M0: compilable skeleton — Kigi 0.1.0 fork surgery

Hard fork of xai-org/grok-build (Apache-2.0) re-targeted as Kigi, an
unofficial Kimi Code CLI community build.

Rename & identity
- 72 xai-*/xai-grok-* crates -> kigi-* (explicit: xai-grok-pager-bin ->
  kigi-bin [binary `kigi`], xai-grok-pager -> kigi-tui; rest mechanical);
  ptyctl, ptyctl-cli, third_party/ unchanged; proto package
  xai.grok.tools.v1 -> kigi.tools.v1
- Config home ~/.kigi (KIGI_SHARE_DIR override), env prefix GROK_* ->
  KIGI_*, `kigi --version` carries the unofficial-community-build notice
- clap identity, help text, startup banner, prompt templates rebranded
  (templates re-encrypted)

Deletions (PRD removal list #5/#6/#7/#9/#10)
- voice input (xai-grok-voice) and all TUI wiring
- telemetry: Mixpanel client, external OTel stream, Sentry, OTLP layers,
  trace/GCS/S3 upload queues (kigi-file-utils halved), workspace upload
  module & dc_log, heap-profile uploader, auth-diagnostics uploader,
  session-analytics halves of feedback; local zero-egress observability
  preserved in new kigi-log crate (unified log, --debug firehose,
  subsystem file logs, opt-in instrumentation)
- announcements (crate, remote-settings fields, TUI surfaces)
- plugin marketplace (crate, sources/browse/CTA/extensions-modal tab);
  direct plugin install/uninstall/update via kigi-agent git_install kept
- relay/gateway/assets endpoints and features (agent relay, headless
  relay transport, gateway bridge, LeaderEnvUrls); leader IPC socket now
  ~/.kigi/leader.sock + KIGI_LEADER_SOCKET, no ws-url derivation
- functional types rehomed instead of deleted: PermissionMode ->
  kigi-config-types, McpInitStrategy -> kigi-mcp, PrCreationSource ->
  session signals, TerminalDiagnostics -> kigi-pager-render, agent_id ->
  shell util

Endpoints
- kigi-env rewritten: single production KigiEndpoints {coding_api_base_url
  https://api.kimi.com/coding/v1 (KIGI_CODE_BASE_URL), oauth_host
  https://auth.kimi.com (KIGI_OAUTH_HOST), update_base_url (GitHub
  Releases API), upgrade_page_url}; GrokBuildEnvironment enum deleted

Toolchain & workspace hygiene
- Rust 1.97.0 pinned; edition 2024; full cargo update; git2 hoisted to
  workspace at 0.21 (Option->Result API migration), quick-xml 0.41
- Root Cargo.toml hand-maintained (PRD §8.1): version 0.1.0 inherited by
  all members, members sorted, unused deps pruned
- cargo-deny advisories gate (deny.toml with documented transitive
  exceptions); CI workflow (check/clippy/fmt/deny/test, macOS+Linux)
- cross-crate test seams re-gated behind `test-support` cargo feature;
  insta snapshot baselines renamed to the kigi_tui prefix
- clippy --workspace --all-targets: zero warnings; fmt clean

Fixes surfaced by the port
- updater probe/installer divergence (bin/kigi vs bin/grok symlink set)
- idle model-metadata refresh dead under KIGI_CODE_BASE_URL override
  (new is_effective_coding_endpoint_url, loopback+override aware)
- macOS symlinked-TMPDIR fixture canonicalization (foreign_sessions,
  fast-worktree); RSS measurement tests serialized via serial_test

Docs & legal (Apache §4)
- NOTICE added (upstream attribution + change statement); THIRD-PARTY
  notices sustained; kigi-tools ported-code notices extended; README,
  CONTRIBUTING, SECURITY, AGENTS.md rewritten

Out of scope for M0 (tracked): Kimi auth/inference (M1), search/fetch,
command parity, config import (M2), Computer Hub excision & final
brand-token sweep (M2), distribution & self-update rewrite (M3).

crates/codegen/kigi-codebase-graph/src/bin/bench_file_listing.rs

@@ -0,0 +1,233 @@
+//! Benchmark for comparing git CLI vs git2 file listing.
+//!
+//! Usage: cargo run --bin bench_file_listing --release -- [path] [cli|git2|git2-index|both]
+
+use std::path::Path;
+use std::process::Command;
+use std::time::Instant;
+
+use git2::{Repository, StatusOptions};
+use kigi_codebase_graph::LanguageRegistry;
+
+fn main() {
+    let args: Vec<String> = std::env::args().collect();
+    let path_str = if let Some(p) = args.get(1) {
+        p.clone()
+    } else if let Ok(p) = std::env::var("BENCH_REPO_ROOT").or_else(|_| std::env::var("XAI_ROOT")) {
+        p
+    } else {
+        eprintln!("Usage: bench_file_listing <path> [cli|git2|git2-index|both]");
+        eprintln!("Or set BENCH_REPO_ROOT to a large checkout to bench against");
+        std::process::exit(1);
+    };
+    let mode = args.get(2).map(|s| s.as_str()).unwrap_or("both");
+
+    let root_path = Path::new(&path_str);
+    let registry = LanguageRegistry::new();
+
+    match mode {
+        "cli" => {
+            let start = Instant::now();
+            let files = collect_files_cli(root_path, &registry);
+            let elapsed = start.elapsed();
+            println!("CLI: {} files in {:?}", files.len(), elapsed);
+        }
+        "git2" => {
+            let start = Instant::now();
+            let files = collect_files_git2(root_path, &registry);
+            let elapsed = start.elapsed();
+            println!("git2: {} files in {:?}", files.len(), elapsed);
+        }
+        "git2-index" => {
+            let start = Instant::now();
+            let files = collect_files_git2_index_only(root_path, &registry);
+            let elapsed = start.elapsed();
+            println!("git2 (index only): {} files in {:?}", files.len(), elapsed);
+        }
+        _ => {
+            // Run all three methods multiple times for comparison
+            println!("Benchmarking file listing for: {}", root_path.display());
+            println!();
+
+            let iterations = 5;
+
+            // Warm up
+            let _ = collect_files_cli(root_path, &registry);
+            let _ = collect_files_git2(root_path, &registry);
+            let _ = collect_files_git2_index_only(root_path, &registry);
+
+            // CLI benchmark
+            let mut cli_times = Vec::with_capacity(iterations);
+            let mut cli_count = 0;
+            for _ in 0..iterations {
+                let start = Instant::now();
+                let files = collect_files_cli(root_path, &registry);
+                cli_times.push(start.elapsed());
+                cli_count = files.len();
+            }
+
+            // git2 benchmark (with untracked)
+            let mut git2_times = Vec::with_capacity(iterations);
+            let mut git2_count = 0;
+            for _ in 0..iterations {
+                let start = Instant::now();
+                let files = collect_files_git2(root_path, &registry);
+                git2_times.push(start.elapsed());
+                git2_count = files.len();
+            }
+
+            // git2 index-only benchmark
+            let mut git2_index_times = Vec::with_capacity(iterations);
+            let mut git2_index_count = 0;
+            for _ in 0..iterations {
+                let start = Instant::now();
+                let files = collect_files_git2_index_only(root_path, &registry);
+                git2_index_times.push(start.elapsed());
+                git2_index_count = files.len();
+            }
+
+            // Print results
+            let cli_avg = cli_times.iter().sum::<std::time::Duration>() / iterations as u32;
+            let git2_avg = git2_times.iter().sum::<std::time::Duration>() / iterations as u32;
+            let git2_index_avg =
+                git2_index_times.iter().sum::<std::time::Duration>() / iterations as u32;
+
+            println!("Results ({} iterations):", iterations);
+            println!(
+                "  CLI:               {} files, avg {:?}",
+                cli_count, cli_avg
+            );
+            println!(
+                "  git2 (+ untracked): {} files, avg {:?}",
+                git2_count, git2_avg
+            );
+            println!(
+                "  git2 (index only):  {} files, avg {:?}",
+                git2_index_count, git2_index_avg
+            );
+            println!();
+
+            let speedup = cli_avg.as_secs_f64() / git2_index_avg.as_secs_f64();
+            if speedup > 1.0 {
+                println!("git2 (index only) is {:.2}x faster than CLI", speedup);
+            } else {
+                println!("CLI is {:.2}x faster than git2 (index only)", 1.0 / speedup);
+            }
+        }
+    }
+}
+
+/// Collect files using git CLI (original approach)
+fn collect_files_cli(root_path: &Path, registry: &LanguageRegistry) -> Vec<std::path::PathBuf> {
+    // Get tracked files
+    let tracked_output = Command::new("git")
+        .args(["ls-files"])
+        .current_dir(root_path)
+        .output();
+
+    let tracked_output = match tracked_output {
+        Ok(o) if o.status.success() => o,
+        _ => return vec![],
+    };
+
+    // Get untracked files
+    let untracked_output = Command::new("git")
+        .args(["ls-files", "--others", "--exclude-standard"])
+        .current_dir(root_path)
+        .output()
+        .ok();
+
+    let tracked_str = String::from_utf8_lossy(&tracked_output.stdout);
+    let mut files: Vec<std::path::PathBuf> = tracked_str
+        .lines()
+        .filter(|line| registry.is_supported(Path::new(line)))
+        .map(|line| root_path.join(line))
+        .collect();
+
+    if let Some(output) = untracked_output
+        && output.status.success()
+    {
+        let untracked_str = String::from_utf8_lossy(&output.stdout);
+        let untracked_files: Vec<std::path::PathBuf> = untracked_str
+            .lines()
+            .filter(|line| registry.is_supported(Path::new(line)))
+            .map(|line| root_path.join(line))
+            .collect();
+        files.extend(untracked_files);
+    }
+
+    files
+}
+
+/// Collect files using git2 (new approach)
+fn collect_files_git2(root_path: &Path, registry: &LanguageRegistry) -> Vec<std::path::PathBuf> {
+    let repo = match Repository::open(root_path) {
+        Ok(r) => r,
+        Err(_) => return vec![],
+    };
+
+    let index = match repo.index() {
+        Ok(i) => i,
+        Err(_) => return vec![],
+    };
+
+    let mut files: Vec<std::path::PathBuf> = index
+        .iter()
+        .filter_map(|entry| {
+            let path_str = std::str::from_utf8(&entry.path).ok()?;
+            if registry.is_supported(Path::new(path_str)) {
+                Some(root_path.join(path_str))
+            } else {
+                None
+            }
+        })
+        .collect();
+
+    // Get untracked files
+    let mut status_opts = StatusOptions::new();
+    status_opts
+        .include_untracked(true)
+        .recurse_untracked_dirs(true)
+        .exclude_submodules(true);
+
+    if let Ok(statuses) = repo.statuses(Some(&mut status_opts)) {
+        for status_entry in statuses.iter() {
+            if status_entry.status().is_wt_new()
+                && let Ok(path_str) = status_entry.path()
+                && registry.is_supported(Path::new(path_str))
+            {
+                files.push(root_path.join(path_str));
+            }
+        }
+    }
+
+    files
+}
+
+/// Collect files using git2 index only (tracked files only, no untracked)
+fn collect_files_git2_index_only(
+    root_path: &Path,
+    registry: &LanguageRegistry,
+) -> Vec<std::path::PathBuf> {
+    let repo = match Repository::open(root_path) {
+        Ok(r) => r,
+        Err(_) => return vec![],
+    };
+
+    let index = match repo.index() {
+        Ok(i) => i,
+        Err(_) => return vec![],
+    };
+
+    index
+        .iter()
+        .filter_map(|entry| {
+            let path_str = std::str::from_utf8(&entry.path).ok()?;
+            if registry.is_supported(Path::new(path_str)) {
+                Some(root_path.join(path_str))
+            } else {
+                None
+            }
+        })
+        .collect()
+}

crates/codegen/kigi-codebase-graph/src/bin/bench_index.rs

@@ -0,0 +1,64 @@
+//! Benchmark binary for index building.
+
+use std::path::Path;
+use std::time::Instant;
+
+use kigi_codebase_graph::{IndexBuilder, LanguageRegistry};
+
+// Use mimalloc for faster allocation in multi-threaded workloads
+#[global_allocator]
+static GLOBAL: mimalloc::MiMalloc = mimalloc::MiMalloc;
+
+fn main() {
+    let args: Vec<String> = std::env::args().collect();
+    let path = if let Some(p) = args.get(1) {
+        p.clone()
+    } else if let Ok(p) = std::env::var("BENCH_REPO_ROOT").or_else(|_| std::env::var("XAI_ROOT")) {
+        p
+    } else {
+        eprintln!("Usage: bench_index <path>");
+        eprintln!("Or set BENCH_REPO_ROOT to a large checkout to bench against");
+        std::process::exit(1);
+    };
+
+    // First, verify all queries compile
+    println!("Verifying query compilation...");
+    let registry = LanguageRegistry::new();
+    for ext in &["ts", "tsx", "js", "jsx", "rs", "go", "py"] {
+        match registry.for_extension(ext) {
+            Some(config) => match config.compile_query() {
+                Ok(query) => {
+                    println!("  .{}: OK ({} patterns)", ext, query.pattern_count());
+                }
+                Err(e) => {
+                    println!("  .{}: FAILED - {:?}", ext, e);
+                }
+            },
+            None => println!("  .{}: NOT SUPPORTED", ext),
+        }
+    }
+    println!();
+
+    let root_path = Path::new(&path);
+
+    println!("Building index for: {}", root_path.display());
+    let start = Instant::now();
+
+    let index = IndexBuilder::new()
+        .build(root_path)
+        .expect("Failed to build index");
+
+    let elapsed = start.elapsed();
+    let (file_count, defs, refs) = index.stats();
+
+    println!("Files indexed: {}", file_count);
+    println!(
+        "Indexed {} definitions, {} references in {:?}",
+        defs, refs, elapsed
+    );
+    println!("Aliases: {}", index.alias_count());
+    println!(
+        "Files/sec: {:.0}",
+        file_count as f64 / elapsed.as_secs_f64()
+    );
+}

crates/codegen/kigi-codebase-graph/src/bin/code_graph.rs

@@ -0,0 +1,394 @@
+//! CLI tool for code graph navigation.
+//!
+//! Provides go-to-definition and go-to-references functionality.
+//!
+//! # Usage
+//!
+//! ```bash
+//! # Build the index for a repository
+//! code-graph index /path/to/repo
+//!
+//! # Build the index with custom cache location
+//! code-graph index /path/to/repo --cache /path/to/cache.bin
+//!
+//! # Go to definition (by position)
+//! code-graph definition /path/to/repo --file src/main.rs --row 10 --col 15
+//!
+//! # Go to definition (by symbol name)
+//! code-graph definition /path/to/repo --symbol MyStruct
+//!
+//! # Go to references (by position)
+//! code-graph references /path/to/repo --file src/main.rs --row 10 --col 15
+//!
+//! # Go to references (by symbol name)
+//! code-graph references /path/to/repo --symbol MyStruct
+//!
+//! # Show index statistics
+//! code-graph stats /path/to/repo
+//! ```
+
+use std::path::{Path, PathBuf};
+use std::time::Instant;
+
+use clap::{Parser, Subcommand};
+
+use kigi_codebase_graph::{
+    IndexBuilder, Navigator, ScopeGraphIndex, get_cache_path, load_index, save_index,
+};
+
+// Use mimalloc for faster allocation
+#[global_allocator]
+static GLOBAL: mimalloc::MiMalloc = mimalloc::MiMalloc;
+
+#[derive(Parser)]
+#[command(name = "code-graph")]
+#[command(author, version, about = "High-performance code navigation tool", long_about = None)]
+struct Cli {
+    #[command(subcommand)]
+    command: Commands,
+}
+
+#[derive(Subcommand)]
+enum Commands {
+    /// Build or rebuild the index for a repository
+    Index {
+        /// Path to the repository
+        path: PathBuf,
+        /// Custom cache file path (default: <repo>/.goto_index.bin)
+        #[arg(short, long)]
+        cache: Option<PathBuf>,
+        /// Force rebuild even if cache exists
+        #[arg(short, long)]
+        force: bool,
+        /// Number of threads to use
+        #[arg(short, long)]
+        threads: Option<usize>,
+    },
+
+    /// Go to definition for a symbol
+    Definition {
+        /// Path to the repository
+        path: PathBuf,
+        /// Custom cache file path (default: <repo>/.goto_index.bin)
+        #[arg(long)]
+        cache: Option<PathBuf>,
+        /// File path (for position-based lookup)
+        #[arg(short, long)]
+        file: Option<PathBuf>,
+        /// Row number (1-indexed, for position-based lookup)
+        #[arg(short, long)]
+        row: Option<usize>,
+        /// Column number (1-indexed, for position-based lookup)
+        #[arg(short, long)]
+        col: Option<usize>,
+        /// Symbol name (for direct lookup)
+        #[arg(short, long)]
+        symbol: Option<String>,
+        /// Output as JSON
+        #[arg(long)]
+        json: bool,
+    },
+
+    /// Go to references for a symbol
+    References {
+        /// Path to the repository
+        path: PathBuf,
+        /// Custom cache file path (default: <repo>/.goto_index.bin)
+        #[arg(long)]
+        cache: Option<PathBuf>,
+        /// File path (for position-based lookup)
+        #[arg(short, long)]
+        file: Option<PathBuf>,
+        /// Row number (1-indexed, for position-based lookup)
+        #[arg(short, long)]
+        row: Option<usize>,
+        /// Column number (1-indexed, for position-based lookup)
+        #[arg(short, long)]
+        col: Option<usize>,
+        /// Symbol name (for direct lookup)
+        #[arg(short, long)]
+        symbol: Option<String>,
+        /// Include definition in results
+        #[arg(long)]
+        include_definition: bool,
+        /// Output as JSON
+        #[arg(long)]
+        json: bool,
+    },
+
+    /// Show index statistics
+    Stats {
+        /// Path to the repository
+        path: PathBuf,
+        /// Custom cache file path (default: <repo>/.goto_index.bin)
+        #[arg(long)]
+        cache: Option<PathBuf>,
+    },
+}
+
+fn main() {
+    let cli = Cli::parse();
+
+    match cli.command {
+        Commands::Index {
+            path,
+            cache,
+            force,
+            threads,
+        } => {
+            cmd_index(&path, cache.as_deref(), force, threads);
+        }
+        Commands::Definition {
+            path,
+            cache,
+            file,
+            row,
+            col,
+            symbol,
+            json,
+        } => {
+            cmd_definition(&path, cache.as_deref(), file, row, col, symbol, json);
+        }
+        Commands::References {
+            path,
+            cache,
+            file,
+            row,
+            col,
+            symbol,
+            include_definition,
+            json,
+        } => {
+            cmd_references(
+                &path,
+                cache.as_deref(),
+                file,
+                row,
+                col,
+                symbol,
+                include_definition,
+                json,
+            );
+        }
+        Commands::Stats { path, cache } => {
+            cmd_stats(&path, cache.as_deref());
+        }
+    }
+}
+
+/// Get the effective cache path - use custom if provided, otherwise default.
+fn effective_cache_path(repo_path: &Path, custom_cache: Option<&Path>) -> PathBuf {
+    custom_cache
+        .map(|p| p.to_path_buf())
+        .unwrap_or_else(|| get_cache_path(repo_path))
+}
+
+/// Load index from cache or build if necessary.
+fn load_or_build_index(repo_path: &Path, cache_path: &Path) -> ScopeGraphIndex {
+    if let Ok(index) = load_index(cache_path) {
+        println!("Loaded index from cache: {}", cache_path.display());
+        return index;
+    }
+
+    println!("Building index for: {}", repo_path.display());
+    let start = Instant::now();
+
+    let index = IndexBuilder::new()
+        .build(repo_path)
+        .expect("Failed to build index");
+
+    let elapsed = start.elapsed();
+    let (files, defs, refs) = index.stats();
+    println!(
+        "Built index: {} files, {} defs, {} refs in {:?}",
+        files, defs, refs, elapsed
+    );
+
+    // Save to cache
+    if let Err(e) = save_index(cache_path, &index) {
+        println!("Warning: Failed to save cache: {}", e);
+    } else {
+        println!("Saved cache to: {}", cache_path.display());
+    }
+
+    index
+}
+
+fn cmd_index(path: &Path, custom_cache: Option<&Path>, _force: bool, threads: Option<usize>) {
+    let cache_path = effective_cache_path(path, custom_cache);
+
+    println!("Building index for: {}", path.display());
+    let start = Instant::now();
+
+    let mut builder = IndexBuilder::new();
+    if let Some(t) = threads {
+        builder = builder.with_threads(t);
+    }
+
+    let index = builder.build(path).expect("Failed to build index");
+
+    let elapsed = start.elapsed();
+    let (files, defs, refs) = index.stats();
+
+    println!("Index built successfully!");
+    println!("  Files indexed: {}", files);
+    println!("  Definitions:   {}", defs);
+    println!("  References:    {}", refs);
+    println!("  Aliases:       {}", index.alias_count());
+    println!("  Time:          {:?}", elapsed);
+
+    // Always save when explicitly indexing
+    if let Err(e) = save_index(&cache_path, &index) {
+        println!("Error saving cache: {}", e);
+        std::process::exit(1);
+    } else {
+        println!("  Cache saved:   {}", cache_path.display());
+    }
+}
+
+fn cmd_definition(
+    repo_path: &Path,
+    custom_cache: Option<&Path>,
+    file: Option<PathBuf>,
+    row: Option<usize>,
+    col: Option<usize>,
+    symbol: Option<String>,
+    json: bool,
+) {
+    let cache_path = effective_cache_path(repo_path, custom_cache);
+    let index = load_or_build_index(repo_path, &cache_path);
+    let navigator = Navigator::new(index);
+
+    let result = match (file, row, col, symbol) {
+        // Position-based lookup
+        (Some(file_path), Some(r), Some(c), _) => {
+            let abs_path = if file_path.is_absolute() {
+                file_path
+            } else {
+                repo_path.join(&file_path)
+            };
+
+            match navigator.goto_definition(&abs_path, r, c) {
+                Ok(r) => r,
+                Err(e) => {
+                    println!("Error: {}", e);
+                    std::process::exit(1);
+                }
+            }
+        }
+        // Symbol-based lookup
+        (_, _, _, Some(sym)) => navigator.goto_definition_by_name(&sym, None),
+        _ => {
+            println!("Error: Must provide either --file, --row, --col OR --symbol");
+            std::process::exit(1);
+        }
+    };
+
+    if json {
+        print_json(&result);
+    } else {
+        println!("Symbol: {}", result.symbol);
+        println!("Definitions ({}):", result.locations.len());
+        for loc in &result.locations {
+            println!("  {}:{}", loc.path, loc.line);
+        }
+    }
+}
+
+fn cmd_references(
+    repo_path: &Path,
+    custom_cache: Option<&Path>,
+    file: Option<PathBuf>,
+    row: Option<usize>,
+    col: Option<usize>,
+    symbol: Option<String>,
+    include_definition: bool,
+    json: bool,
+) {
+    let cache_path = effective_cache_path(repo_path, custom_cache);
+    let index = load_or_build_index(repo_path, &cache_path);
+    let navigator = Navigator::new(index);
+
+    let result = match (file, row, col, symbol) {
+        // Position-based lookup
+        (Some(file_path), Some(r), Some(c), _) => {
+            let abs_path = if file_path.is_absolute() {
+                file_path
+            } else {
+                repo_path.join(&file_path)
+            };
+
+            match navigator.goto_references(&abs_path, r, c, include_definition) {
+                Ok(r) => r,
+                Err(e) => {
+                    println!("Error: {}", e);
+                    std::process::exit(1);
+                }
+            }
+        }
+        // Symbol-based lookup
+        (_, _, _, Some(sym)) => navigator.goto_references_by_name(&sym, None, include_definition),
+        _ => {
+            println!("Error: Must provide either --file, --row, --col OR --symbol");
+            std::process::exit(1);
+        }
+    };
+
+    if json {
+        print_json(&result);
+    } else {
+        println!("Symbol: {}", result.symbol);
+        println!("References ({}):", result.locations.len());
+        for loc in &result.locations {
+            if let Some(sym) = &loc.symbol {
+                println!("  {}:{} (as {})", loc.path, loc.line, sym);
+            } else {
+                println!("  {}:{}", loc.path, loc.line);
+            }
+        }
+    }
+}
+
+fn cmd_stats(path: &Path, custom_cache: Option<&Path>) {
+    let cache_path = effective_cache_path(path, custom_cache);
+    let index = load_or_build_index(path, &cache_path);
+    let (files, defs, refs) = index.stats();
+
+    println!("Index Statistics for: {}", path.display());
+    println!("  Cache location: {}", cache_path.display());
+    println!("  Files indexed:  {}", files);
+    println!("  Definitions:    {}", defs);
+    println!("  References:     {}", refs);
+    println!("  Aliases:        {}", index.alias_count());
+
+    // Top symbols by reference count
+    let ref_counts = index.top_referenced_symbols(10);
+
+    println!("\nTop 10 most referenced symbols:");
+    for (name, count) in &ref_counts {
+        println!("  {:6} {}", count, name);
+    }
+}
+
+fn print_json(result: &kigi_codebase_graph::NavigationResult) {
+    use serde_json::json;
+
+    let locations: Vec<_> = result
+        .locations
+        .iter()
+        .map(|loc| {
+            json!({
+                "path": &loc.path,
+                "line": loc.line,
+                "symbol": loc.symbol,
+            })
+        })
+        .collect();
+
+    let output = json!({
+        "symbol": result.symbol,
+        "locations": locations,
+    });
+
+    println!("{}", serde_json::to_string_pretty(&output).unwrap());
+}

crates/codegen/kigi-codebase-graph/src/interner.rs

@@ -0,0 +1,413 @@
+//! Arena-based string interner for memory-efficient string deduplication.
+//!
+//! This module provides a string interner that stores all strings in a single
+//! contiguous buffer, minimizing allocations and improving cache locality.
+//! It uses hash-based lookup for O(1) interning operations.
+//!
+//! # Design
+//!
+//! The interner uses a two-level lookup approach:
+//! 1. **Primary lookup**: HashMap from 64-bit hash -> list of StringIds with that hash
+//! 2. **Collision resolution**: When hashes collide, actual string content is compared
+//!
+//! This gives O(1) average case for both `intern()` and `get_id()` operations.
+//!
+//! # Example
+//!
+//! ```
+//! use kigi_codebase_graph::interner::StringInterner;
+//!
+//! let mut interner = StringInterner::new();
+//!
+//! let id1 = interner.intern("hello");
+//! let id2 = interner.intern("world");
+//! let id3 = interner.intern("hello"); // Returns same id as id1
+//!
+//! assert_eq!(id1, id3);
+//! assert_ne!(id1, id2);
+//! assert_eq!(interner.get(id1), Some("hello"));
+//! ```
+
+use std::hash::{Hash, Hasher};
+
+use hashbrown::HashMap;
+use nohash_hasher::BuildNoHashHasher;
+use rustc_hash::FxHasher;
+use serde::{Deserialize, Serialize};
+use smallvec::SmallVec;
+
+/// Type alias for HashMap with u64 keys that are already hashed.
+/// Uses NoHashHasher since keys don't need re-hashing.
+type U64NoHashMap<V> = HashMap<u64, V, BuildNoHashHasher<u64>>;
+
+/// A compact identifier for an interned string.
+/// Using u32 allows up to 4 billion unique strings.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash, Serialize, Deserialize)]
+pub struct StringId(u32);
+
+impl StringId {
+    /// Create a new StringId from a raw u32 value.
+    #[inline]
+    pub const fn new(id: u32) -> Self {
+        Self(id)
+    }
+
+    /// Get the raw u32 value.
+    #[inline]
+    pub const fn as_u32(self) -> u32 {
+        self.0
+    }
+}
+
+/// Arena-based string interner for efficient string deduplication.
+///
+/// Stores all strings in a single contiguous buffer to minimize allocations
+/// and improve cache locality. Uses a hash-based lookup for O(1) interning.
+///
+/// The interner stores arbitrary byte sequences, supporting paths and strings
+/// that may not be valid UTF-8.
+#[derive(Debug, Clone)]
+pub struct StringInterner {
+    /// Contiguous storage for all interned byte strings
+    arena: Vec<u8>,
+    /// Maps hash -> StringId(s). Most buckets have exactly one entry.
+    /// Using SmallVec<[StringId; 1]> optimizes for the common case of no collisions.
+    /// Uses NoHashHasher since keys are already hashed.
+    lookup: U64NoHashMap<SmallVec<[StringId; 1]>>,
+    /// Maps StringId to (start, len) in arena
+    offsets: Vec<(u32, u16)>,
+}
+
+impl Default for StringInterner {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl StringInterner {
+    /// Create a new empty interner.
+    pub fn new() -> Self {
+        Self {
+            arena: Vec::new(),
+            lookup: U64NoHashMap::default(),
+            offsets: Vec::new(),
+        }
+    }
+
+    /// Create an interner with pre-allocated capacity.
+    ///
+    /// # Arguments
+    /// * `string_bytes` - Estimated total bytes for all strings
+    /// * `num_strings` - Estimated number of unique strings
+    pub fn with_capacity(string_bytes: usize, num_strings: usize) -> Self {
+        Self {
+            arena: Vec::with_capacity(string_bytes),
+            lookup: U64NoHashMap::with_capacity_and_hasher(
+                num_strings,
+                BuildNoHashHasher::default(),
+            ),
+            offsets: Vec::with_capacity(num_strings),
+        }
+    }
+
+    /// Intern a byte string, returning its StringId.
+    /// If the string is already interned, returns the existing id.
+    ///
+    /// # Complexity
+    /// O(1) average case, O(k) worst case where k is the number of
+    /// hash collisions (typically 0 or 1).
+    pub fn intern_bytes(&mut self, s: &[u8]) -> StringId {
+        let hash = Self::hash_bytes(s);
+
+        // Check if already interned
+        if let Some(ids) = self.lookup.get(&hash) {
+            for &id in ids {
+                if self.get_bytes(id) == Some(s) {
+                    return id;
+                }
+            }
+        }
+
+        // Not found, add new
+        let start = self.arena.len() as u32;
+        let len = s.len() as u16;
+
+        self.arena.extend_from_slice(s);
+
+        let id = StringId::new(self.offsets.len() as u32);
+        self.offsets.push((start, len));
+
+        // Add to lookup
+        self.lookup.entry(hash).or_default().push(id);
+
+        id
+    }
+
+    /// Intern a UTF-8 string. Convenience wrapper around `intern_bytes`.
+    #[inline]
+    pub fn intern(&mut self, s: &str) -> StringId {
+        self.intern_bytes(s.as_bytes())
+    }
+
+    /// Get the StringId for a byte string without interning it.
+    /// Returns None if the string is not in the interner.
+    ///
+    /// # Complexity
+    /// O(1) average case.
+    pub fn get_bytes_id(&self, s: &[u8]) -> Option<StringId> {
+        let hash = Self::hash_bytes(s);
+
+        if let Some(ids) = self.lookup.get(&hash) {
+            for &id in ids {
+                if self.get_bytes(id) == Some(s) {
+                    return Some(id);
+                }
+            }
+        }
+        None
+    }
+
+    /// Get the StringId for a UTF-8 string without interning it.
+    #[inline]
+    pub fn get_id(&self, s: &str) -> Option<StringId> {
+        self.get_bytes_id(s.as_bytes())
+    }
+
+    /// Get the raw bytes for a StringId.
+    ///
+    /// # Complexity
+    /// O(1)
+    pub fn get_bytes(&self, id: StringId) -> Option<&[u8]> {
+        let (start, len) = *self.offsets.get(id.0 as usize)?;
+        self.arena
+            .get(start as usize..(start as usize + len as usize))
+    }
+
+    /// Get the string for a StringId, if it's valid UTF-8.
+    ///
+    /// # Complexity
+    /// O(1)
+    pub fn get(&self, id: StringId) -> Option<&str> {
+        self.get_bytes(id).and_then(|b| std::str::from_utf8(b).ok())
+    }
+
+    /// Get the string for a StringId, with lossy UTF-8 conversion.
+    /// Invalid UTF-8 sequences are replaced with the replacement character.
+    pub fn get_lossy(&self, id: StringId) -> Option<std::borrow::Cow<'_, str>> {
+        self.get_bytes(id).map(String::from_utf8_lossy)
+    }
+
+    /// Number of interned strings.
+    #[inline]
+    pub fn len(&self) -> usize {
+        self.offsets.len()
+    }
+
+    /// Check if the interner is empty.
+    #[inline]
+    pub fn is_empty(&self) -> bool {
+        self.offsets.is_empty()
+    }
+
+    /// Total bytes used by the arena.
+    #[inline]
+    pub fn arena_bytes(&self) -> usize {
+        self.arena.len()
+    }
+
+    /// Compute FxHash of a byte slice.
+    #[inline]
+    fn hash_bytes(s: &[u8]) -> u64 {
+        let mut hasher = FxHasher::default();
+        s.hash(&mut hasher);
+        hasher.finish()
+    }
+
+    /// Iterate over all strings with their IDs (only valid UTF-8).
+    pub fn iter(&self) -> impl Iterator<Item = (StringId, &str)> {
+        self.offsets
+            .iter()
+            .enumerate()
+            .filter_map(|(idx, &(start, len))| {
+                let bytes = self
+                    .arena
+                    .get(start as usize..(start as usize + len as usize))?;
+                let s = std::str::from_utf8(bytes).ok()?;
+                Some((StringId::new(idx as u32), s))
+            })
+    }
+
+    /// Iterate over all byte strings with their IDs.
+    pub fn iter_bytes(&self) -> impl Iterator<Item = (StringId, &[u8])> {
+        self.offsets
+            .iter()
+            .enumerate()
+            .filter_map(|(idx, &(start, len))| {
+                let bytes = self
+                    .arena
+                    .get(start as usize..(start as usize + len as usize))?;
+                Some((StringId::new(idx as u32), bytes))
+            })
+    }
+
+    /// Clear the interner, removing all strings but keeping allocated capacity.
+    pub fn clear(&mut self) {
+        self.arena.clear();
+        self.lookup.clear();
+        self.offsets.clear();
+    }
+
+    /// Get the internal arena for serialization purposes.
+    pub fn arena(&self) -> &[u8] {
+        &self.arena
+    }
+
+    /// Get the internal offsets for serialization purposes.
+    pub fn offsets(&self) -> &[(u32, u16)] {
+        &self.offsets
+    }
+
+    /// Release over-allocated capacity in the arena and offsets buffers.
+    ///
+    /// After a bulk build the arena and offsets Vecs may hold up to 2× their
+    /// actual content due to doubling growth.  Calling this reclaims that
+    /// wasted heap.  The lookup table is intentionally left unshrunk because
+    /// it benefits from load-factor headroom.
+    ///
+    /// This is an internal maintenance hook called by `ScopeGraphIndex::compact()`.
+    pub(crate) fn shrink_to_fit(&mut self) {
+        self.arena.shrink_to_fit();
+        self.offsets.shrink_to_fit();
+    }
+
+    /// Reconstruct an interner from serialized data.
+    ///
+    /// This rebuilds the lookup table from the arena and offsets.
+    pub fn from_parts(arena: Vec<u8>, offsets: Vec<(u32, u16)>) -> Self {
+        let mut lookup: U64NoHashMap<SmallVec<[StringId; 1]>> =
+            U64NoHashMap::with_capacity_and_hasher(offsets.len(), BuildNoHashHasher::default());
+
+        for (idx, &(start, len)) in offsets.iter().enumerate() {
+            if let Some(bytes) = arena.get(start as usize..(start as usize + len as usize)) {
+                let hash = Self::hash_bytes(bytes);
+                let id = StringId::new(idx as u32);
+                lookup.entry(hash).or_default().push(id);
+            }
+        }
+
+        Self {
+            arena,
+            lookup,
+            offsets,
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_basic_interning() {
+        let mut interner = StringInterner::new();
+
+        let id1 = interner.intern("src");
+        let id2 = interner.intern("lib");
+        let id3 = interner.intern("src"); // duplicate
+
+        assert_eq!(id1, id3);
+        assert_ne!(id1, id2);
+        assert_eq!(interner.get(id1), Some("src"));
+        assert_eq!(interner.get(id2), Some("lib"));
+        assert_eq!(interner.len(), 2);
+    }
+
+    #[test]
+    fn test_get_id() {
+        let mut interner = StringInterner::new();
+
+        let id_src = interner.intern("src");
+        let id_lib = interner.intern("lib");
+
+        assert_eq!(interner.get_id("src"), Some(id_src));
+        assert_eq!(interner.get_id("lib"), Some(id_lib));
+        assert_eq!(interner.get_id("nonexistent"), None);
+
+        // get_id should not modify the interner
+        assert_eq!(interner.len(), 2);
+    }
+
+    #[test]
+    fn test_bytes_interning() {
+        let mut interner = StringInterner::new();
+
+        // Valid UTF-8
+        let id1 = interner.intern_bytes(b"hello");
+        assert_eq!(interner.get(id1), Some("hello"));
+
+        // Invalid UTF-8
+        let invalid_utf8: &[u8] = &[0x80, 0x81, 0x82];
+        let id2 = interner.intern_bytes(invalid_utf8);
+        assert_eq!(interner.get(id2), None); // Not valid UTF-8
+        assert_eq!(interner.get_bytes(id2), Some(invalid_utf8));
+
+        // Duplicate bytes return same ID
+        let id3 = interner.intern_bytes(invalid_utf8);
+        assert_eq!(id2, id3);
+    }
+
+    #[test]
+    fn test_many_strings() {
+        let mut interner = StringInterner::new();
+
+        let count = 10_000;
+        let mut ids = Vec::with_capacity(count);
+
+        for i in 0..count {
+            let s = format!("string_{}", i);
+            ids.push(interner.intern(&s));
+        }
+
+        assert_eq!(interner.len(), count);
+
+        // Verify all strings can be looked up
+        for (i, &id) in ids.iter().enumerate() {
+            let s = format!("string_{}", i);
+            assert_eq!(interner.get_id(&s), Some(id));
+            assert_eq!(interner.get(id), Some(s.as_str()));
+        }
+    }
+
+    #[test]
+    fn test_from_parts() {
+        let mut interner = StringInterner::new();
+        interner.intern("hello");
+        interner.intern("world");
+        interner.intern("foo");
+
+        let arena = interner.arena().to_vec();
+        let offsets = interner.offsets().to_vec();
+
+        let restored = StringInterner::from_parts(arena, offsets);
+
+        assert_eq!(restored.len(), 3);
+        assert_eq!(restored.get_id("hello"), Some(StringId::new(0)));
+        assert_eq!(restored.get_id("world"), Some(StringId::new(1)));
+        assert_eq!(restored.get_id("foo"), Some(StringId::new(2)));
+    }
+
+    #[test]
+    fn test_clear() {
+        let mut interner = StringInterner::new();
+        interner.intern("hello");
+        interner.intern("world");
+
+        assert_eq!(interner.len(), 2);
+
+        interner.clear();
+
+        assert_eq!(interner.len(), 0);
+        assert!(interner.is_empty());
+        assert_eq!(interner.get_id("hello"), None);
+    }
+}

crates/codegen/kigi-codebase-graph/src/languages/golang.rs

@@ -0,0 +1,73 @@
+use crate::languages::types::TSLanguageConfig;
+
+pub fn golang() -> TSLanguageConfig {
+    TSLanguageConfig::new(
+        vec!["Go".to_owned(), "go".to_owned()],
+        vec!["go".to_owned()],
+        vec![vec![
+            "function".to_owned(),
+            "type".to_owned(),
+            "struct".to_owned(),
+            "interface".to_owned(),
+            "const".to_owned(),
+            "var".to_owned(),
+            "package".to_owned(),
+        ]],
+        r#"
+        ; Function definitions
+        (function_declaration
+            name: (identifier) @name.definition.function) @definition.function
+        
+        ; Method definitions
+        (method_declaration
+            name: (field_identifier) @name.definition.method) @definition.method
+        
+        ; Type definitions (struct, interface, etc.)
+        (type_declaration
+            (type_spec
+                name: (type_identifier) @name.definition.type)) @definition.type
+        
+        ; Const declarations
+        (const_declaration
+            (const_spec
+                name: (identifier) @name.definition.const)) @definition.const
+        
+        ; Var declarations
+        (var_declaration
+            (var_spec
+                name: (identifier) @name.definition.var)) @definition.var
+        
+        ; ============ REFERENCES ============
+        
+        ; Function calls
+        (call_expression
+            function: (identifier) @name.reference.call) @reference.call
+        
+        ; Method calls
+        (call_expression
+            function: (selector_expression
+                field: (field_identifier) @name.reference.call)) @reference.call
+        
+        ; Type references
+        (type_identifier) @name.reference.type
+        
+        ; Package references in qualified names
+        (qualified_type
+            package: (package_identifier) @name.reference.package
+            name: (type_identifier) @name.reference.type)
+        
+        ; ============ IMPORTS ============
+        
+        ; import "package"
+        (import_spec
+            path: (interpreted_string_literal) @name.reference.import)
+        
+        ; import alias "package"
+        (import_spec
+            name: (package_identifier) @alias.name
+            path: (interpreted_string_literal) @alias.original)
+        "#
+        .to_owned(),
+        || tree_sitter_go::LANGUAGE.into(),
+    )
+}

crates/codegen/kigi-codebase-graph/src/languages/javascript.rs

@@ -0,0 +1,94 @@
+//! JavaScript/JSX language configuration.
+
+use crate::languages::types::TSLanguageConfig;
+
+pub fn js_lang() -> TSLanguageConfig {
+    TSLanguageConfig::new(
+        vec![
+            "JavaScript".to_owned(),
+            "javascript".to_owned(),
+            "js".to_owned(),
+            "jsx".to_owned(),
+        ],
+        vec!["js".to_owned(), "jsx".to_owned()],
+        vec![vec![
+            "function".to_owned(),
+            "class".to_owned(),
+            "variable".to_owned(),
+            "const".to_owned(),
+            "let".to_owned(),
+        ]],
+        r#"
+        ; Class definitions
+        (class_declaration
+            name: (identifier) @name.definition.class) @definition.class
+        
+        ; Function definitions
+        (function_declaration
+            name: (identifier) @name.definition.function) @definition.function
+        
+        ; Arrow function with variable
+        (lexical_declaration
+            (variable_declarator
+                name: (identifier) @name.definition.function
+                value: (arrow_function))) @definition.function
+        
+        ; Method definitions
+        (method_definition
+            name: (property_identifier) @name.definition.method) @definition.method
+        
+        ; Variable declarations
+        (lexical_declaration
+            (variable_declarator
+                name: (identifier) @name.definition.variable)) @definition.variable
+        
+        ; Var declarations
+        (variable_declaration
+            (variable_declarator
+                name: (identifier) @name.definition.variable)) @definition.variable
+        
+        ; ============ REFERENCES ============
+        
+        ; Function calls
+        (call_expression
+            function: (identifier) @name.reference.call) @reference.call
+        
+        ; Method calls
+        (call_expression
+            function: (member_expression
+                property: (property_identifier) @name.reference.call)) @reference.call
+        
+        ; JSX element names
+        (jsx_opening_element
+            name: (identifier) @name.reference.jsx)
+        
+        (jsx_self_closing_element
+            name: (identifier) @name.reference.jsx)
+        
+        ; ============ IMPORTS ============
+        
+        ; Named imports: import { Foo } from 'bar'
+        (import_specifier
+            name: (identifier) @name.reference.import)
+        
+        ; Default import: import Foo from 'bar'
+        (import_clause
+            (identifier) @name.reference.import)
+        
+        ; Import alias: import { Foo as Bar } from 'bar'
+        (import_specifier
+            name: (identifier) @alias.original
+            alias: (identifier) @alias.name)
+        
+        ; Named exports: export { Foo }
+        (export_specifier
+            name: (identifier) @name.reference.export)
+        
+        ; Array element identifiers: [foo, bar] (e.g., React useCallback/useEffect dependency arrays)
+        (array
+            (identifier) @name.reference.variable)
+        "#
+        .to_owned(),
+        || tree_sitter_javascript::LANGUAGE.into(),
+    )
+}

crates/codegen/kigi-codebase-graph/src/languages/mod.rs

@@ -0,0 +1,146 @@
+mod golang;
+mod javascript;
+mod python;
+mod rust;
+mod ts;
+pub mod types;
+
+use std::collections::HashMap;
+use std::hash::{Hash, Hasher};
+use std::path::Path;
+use std::sync::Arc;
+
+pub use golang::golang;
+pub use javascript::js_lang;
+pub use python::python_lang;
+pub use rust::rust_lang;
+pub use ts::ts_lang;
+pub use types::TSLanguageConfig;
+
+/// Registry of all supported languages.
+///
+/// Provides lookup by extension and language ID, and can check if two
+/// extensions belong to the same language family.
+pub struct LanguageRegistry {
+    /// All registered language configs.
+    configs: Vec<Arc<TSLanguageConfig>>,
+    /// Mapping from file extension to language config.
+    by_extension: HashMap<String, Arc<TSLanguageConfig>>,
+    /// Mapping from language ID to language config.
+    by_id: HashMap<String, Arc<TSLanguageConfig>>,
+}
+
+impl LanguageRegistry {
+    /// Create a new language registry with all supported languages.
+    pub fn new() -> Self {
+        let configs: Vec<Arc<TSLanguageConfig>> = vec![
+            Arc::new(rust_lang()),
+            Arc::new(ts_lang()),
+            Arc::new(js_lang()),
+            Arc::new(golang()),
+            Arc::new(python_lang()),
+        ];
+
+        let mut by_extension = HashMap::new();
+        let mut by_id = HashMap::new();
+
+        for config in &configs {
+            for ext in config.file_extensions() {
+                by_extension.insert(ext.clone(), Arc::clone(config));
+            }
+            for id in config.language_ids() {
+                by_id.insert(id.clone(), Arc::clone(config));
+            }
+        }
+
+        Self {
+            configs,
+            by_extension,
+            by_id,
+        }
+    }
+
+    /// Get a language config by file extension.
+    pub fn for_extension(&self, ext: &str) -> Option<Arc<TSLanguageConfig>> {
+        self.by_extension.get(ext).cloned()
+    }
+
+    /// Get a language config by language ID.
+    pub fn for_id(&self, id: &str) -> Option<Arc<TSLanguageConfig>> {
+        self.by_id.get(id).cloned()
+    }
+
+    /// Get a language config for a file path.
+    ///
+    /// Extracts the extension from the path and looks up the config.
+    pub fn for_file_path(&self, path: impl AsRef<Path>) -> Option<Arc<TSLanguageConfig>> {
+        let path = path.as_ref();
+        let ext = path.extension()?.to_str()?;
+        self.for_extension(ext)
+    }
+
+    /// Check if a file path is supported (has a supported extension).
+    pub fn is_supported(&self, path: impl AsRef<Path>) -> bool {
+        self.for_file_path(path).is_some()
+    }
+
+    /// Get all supported file extensions.
+    pub fn supported_extensions(&self) -> Vec<&str> {
+        self.by_extension.keys().map(|s| s.as_str()).collect()
+    }
+
+    /// Get all language configs.
+    pub fn all_configs(&self) -> &[Arc<TSLanguageConfig>] {
+        &self.configs
+    }
+
+    /// Check if two file extensions belong to the same language.
+    ///
+    /// Returns true if both extensions are registered under the same language config.
+    pub fn extensions_same_language(&self, ext1: &str, ext2: &str) -> bool {
+        if ext1 == ext2 {
+            return true;
+        }
+
+        match (self.by_extension.get(ext1), self.by_extension.get(ext2)) {
+            (Some(config1), Some(config2)) => {
+                // Compare by primary language ID (they point to the same config)
+                config1.primary_language_id() == config2.primary_language_id()
+            }
+            _ => false,
+        }
+    }
+
+    /// Compute a hash of all tree-sitter queries across all languages.
+    ///
+    /// This is used to detect when queries change, which should trigger
+    /// a rebuild of the index even if file contents haven't changed.
+    ///
+    /// The hash is computed by:
+    /// 1. Sorting languages by their primary ID for deterministic ordering
+    /// 2. Hashing each language's query string in order
+    /// 3. Combining into a single u64 hash
+    pub fn compute_query_hash(&self) -> u64 {
+        use std::collections::hash_map::DefaultHasher;
+
+        // Sort configs by primary language ID for deterministic ordering
+        let mut sorted_configs: Vec<_> = self.configs.iter().collect();
+        sorted_configs.sort_by_key(|c| c.primary_language_id());
+
+        let mut hasher = DefaultHasher::new();
+
+        for config in sorted_configs {
+            // Hash the language ID and query together
+            config.primary_language_id().hash(&mut hasher);
+            config.file_definition_queries().hash(&mut hasher);
+        }
+
+        hasher.finish()
+    }
+}
+
+impl Default for LanguageRegistry {
+    fn default() -> Self {
+        Self::new()
+    }
+}

crates/codegen/kigi-codebase-graph/src/languages/python.rs

@@ -0,0 +1,38 @@
+//! Python language configuration.
+
+use crate::languages::types::TSLanguageConfig;
+
+pub fn python_lang() -> TSLanguageConfig {
+    TSLanguageConfig::new(
+        vec!["Python".to_owned(), "python".to_owned(), "py".to_owned()],
+        vec!["py".to_owned()],
+        vec![vec![
+            "function".to_owned(),
+            "class".to_owned(),
+            "variable".to_owned(),
+            "module".to_owned(),
+        ]],
+        // Python definitions query
+        r#"
+        ; Class definitions
+        (class_definition
+            name: (identifier) @name.definition.class) @definition.class
+        
+        ; Function definitions
+        (function_definition
+            name: (identifier) @name.definition.function) @definition.function
+        
+        ; ============ REFERENCES ============
+        
+        ; Function calls (direct and method calls)
+        (call
+            function: [
+                (identifier) @name.reference.call
+                (attribute
+                    attribute: (identifier) @name.reference.call)
+            ]) @reference.call
+        "#
+        .to_owned(),
+        || tree_sitter_python::LANGUAGE.into(),
+    )
+}

crates/codegen/kigi-codebase-graph/src/languages/rust.rs

@@ -0,0 +1,198 @@
+use crate::languages::types::TSLanguageConfig;
+
+pub fn rust_lang() -> TSLanguageConfig {
+    TSLanguageConfig::new(
+        vec!["Rust".to_owned(), "rust".to_owned(), "rs".to_owned()],
+        vec!["rs".to_owned()],
+        vec![vec![
+            "const".to_owned(),
+            "function".to_owned(),
+            "variable".to_owned(),
+            "struct".to_owned(),
+            "enum".to_owned(),
+            "union".to_owned(),
+            "typedef".to_owned(),
+            "interface".to_owned(),
+            "field".to_owned(),
+            "enumerator".to_owned(),
+            "module".to_owned(),
+            "label".to_owned(),
+            "lifetime".to_owned(),
+        ]],
+        r#"; ADT definitions
+
+        (struct_item
+            name: (type_identifier) @name.definition.class) @definition.class
+        
+        (enum_item
+            name: (type_identifier) @name.definition.class) @definition.class
+        
+        (union_item
+            name: (type_identifier) @name.definition.class) @definition.class
+        
+        ; type aliases
+        
+        (type_item
+            name: (type_identifier) @name.definition.class) @definition.class
+        
+        ; method definitions
+        
+        (declaration_list
+            (function_item
+                name: (identifier) @name.definition.method)) @definition.method
+        
+        ; function definitions
+        
+        (function_item
+            name: (identifier) @name.definition.function) @definition.function
+        
+        ; trait definitions
+        (trait_item
+            name: (type_identifier) @name.definition.interface) @definition.interface
+        
+        ; module definitions
+        (mod_item
+            name: (identifier) @name.definition.module) @definition.module
+        
+        ; macro definitions
+        
+        (macro_definition
+            name: (identifier) @name.definition.macro) @definition.macro
+        
+        ; const and static definitions
+        (const_item
+            name: (identifier) @name.definition.variable) @definition.variable
+        
+        (static_item
+            name: (identifier) @name.definition.variable) @definition.variable
+        
+        ; ============ REFERENCES ============
+        
+        ; Function and method calls
+        (call_expression
+            function: (identifier) @name.reference.call) @reference.call
+        
+        (call_expression
+            function: (field_expression
+                field: (field_identifier) @name.reference.call)) @reference.call
+        
+        (macro_invocation
+            macro: (identifier) @name.reference.call) @reference.call
+        
+        ; implementations
+        
+        (impl_item
+            trait: (type_identifier) @name.reference.implementation) @reference.implementation
+        
+        (impl_item
+            type: (type_identifier) @name.reference.implementation
+            !trait) @reference.implementation
+        
+        ; ============ USE/IMPORT REFERENCES ============
+        
+        ; Simple use: use Foo;
+        (use_declaration
+            argument: (identifier) @name.reference.import) @reference.import
+        
+        ; Scoped use: use foo::Bar;
+        (use_declaration
+            argument: (scoped_identifier
+                name: (identifier) @name.reference.import)) @reference.import
+        
+        ; Use list: use foo::{Bar, Baz};
+        (use_declaration
+            argument: (scoped_use_list
+                list: (use_list
+                    (identifier) @name.reference.import)))
+        
+        ; Nested scoped use: use foo::bar::{Baz, Qux};
+        (use_declaration
+            argument: (scoped_use_list
+                list: (use_list
+                    (scoped_identifier
+                        name: (identifier) @name.reference.import))))
+        
+        ; Use with alias: use Foo as Bar;
+        (use_declaration
+            argument: (use_as_clause
+                path: (identifier) @name.reference.import))
+        
+        (use_declaration
+            argument: (use_as_clause
+                path: (scoped_identifier
+                    name: (identifier) @name.reference.import)))
+        
+        ; ============ ALIAS TRACKING ============
+        ; These patterns capture alias relationships for unified lookups
+        
+        ; use Foo as Bar - captures original and alias
+        (use_declaration
+            argument: (use_as_clause
+                path: (identifier) @alias.original
+                alias: (identifier) @alias.name))
+        
+        ; use foo::Bar as Baz - scoped version
+        (use_declaration
+            argument: (use_as_clause
+                path: (scoped_identifier
+                    name: (identifier) @alias.original)
+                alias: (identifier) @alias.name))
+        
+        ; ============ TYPE REFERENCES ============
+        
+        ; Type identifiers in function parameters
+        (parameter
+            type: (type_identifier) @name.reference.type)
+        
+        ; Return types
+        (function_item
+            return_type: (type_identifier) @name.reference.type)
+        
+        ; Struct fields
+        (field_declaration
+            type: (type_identifier) @name.reference.type)
+        
+        ; Let bindings with type annotation
+        (let_declaration
+            type: (type_identifier) @name.reference.type)
+        
+        ; Generic type arguments: Vec<Foo>
+        (type_arguments
+            (type_identifier) @name.reference.type)
+        
+        ; Scoped type identifier: foo::Bar
+        (scoped_type_identifier
+            name: (type_identifier) @name.reference.type)
+        
+        ; Reference types: &Foo
+        (reference_type
+            type: (type_identifier) @name.reference.type)
+        
+        ; Tuple struct patterns
+        (tuple_struct_pattern
+            type: (identifier) @name.reference.type)
+        
+        ; Struct expressions: Foo { ... }
+        (struct_expression
+            name: (type_identifier) @name.reference.type)
+        
+        ; Tuple struct expressions: Foo(...)
+        (call_expression
+            function: (scoped_identifier
+                name: (identifier) @name.reference.call))
+        
+        ; Path segments in scoped identifiers: foo::Bar::baz()
+        ; This captures types used in paths like SomeType::method()
+        (scoped_identifier
+            path: (scoped_identifier
+                name: (identifier) @name.reference.type))
+        
+        ; Direct scoped calls with type in path: Foo::bar()
+        (scoped_identifier
+            path: (identifier) @name.reference.type
+            name: (identifier))
+        "#
+        .to_owned(),
+        || tree_sitter_rust::LANGUAGE.into(),
+    )
+}

crates/codegen/kigi-codebase-graph/src/languages/ts.rs

@@ -0,0 +1,241 @@
+use crate::languages::types::TSLanguageConfig;
+
+pub fn ts_lang() -> TSLanguageConfig {
+    TSLanguageConfig::new(
+        vec![
+            "Typescript".to_owned(),
+            "TSX".to_owned(),
+            "typescript".to_owned(),
+            "tsx".to_owned(),
+        ],
+        vec!["ts".to_owned(), "tsx".to_owned()],
+        vec![vec![
+            "function".to_owned(),
+            "class".to_owned(),
+            "interface".to_owned(),
+            "type".to_owned(),
+            "enum".to_owned(),
+            "variable".to_owned(),
+            "const".to_owned(),
+            "let".to_owned(),
+        ]],
+        // Comprehensive TypeScript query with full type coverage
+        r#"
+          ;; === DEFINITIONS ===
+          
+          (function_signature
+            name: (identifier) @name.definition.function) @definition.function
+          
+          (method_signature
+            name: (property_identifier) @name.definition.method) @definition.method
+          
+          (abstract_method_signature
+            name: (property_identifier) @name.definition.method) @definition.method
+          
+          (abstract_class_declaration
+            name: (type_identifier) @name.definition.class) @definition.class
+          
+          (module
+            name: (identifier) @name.definition.module) @definition.module
+          
+          (interface_declaration
+            name: (type_identifier) @name.definition.interface) @definition.interface
+          
+          (function_declaration
+            name: (identifier) @name.definition.function) @definition.function
+          
+          (method_definition
+            name: (property_identifier) @name.definition.method) @definition.method
+          
+          (class_declaration
+            name: (type_identifier) @name.definition.class) @definition.class
+          
+          (type_alias_declaration
+            name: (type_identifier) @name.definition.type) @definition.type
+          
+          (enum_declaration
+            name: (identifier) @name.definition.enum) @definition.enum
+          
+          ;; Arrow function assigned to variable: const foo = () => {}
+          (lexical_declaration
+            (variable_declarator
+              name: (identifier) @name.definition.function
+              value: (arrow_function))) @definition.function
+          
+          ;; React component patterns: const Foo = React.forwardRef(...), React.memo(...)
+          (lexical_declaration
+            (variable_declarator
+              name: (identifier) @name.definition.function
+              value: (call_expression))) @definition.function
+          
+          ;; Variable declarations (const/let)
+          (lexical_declaration
+            (variable_declarator
+              name: (identifier) @name.definition.variable)) @definition.variable
+          
+          ;; Var declarations
+          (variable_declaration
+            (variable_declarator
+              name: (identifier) @name.definition.variable)) @definition.variable
+          
+          ;; Exported variable declarations: export const foo = ...
+          (export_statement
+            (lexical_declaration
+              (variable_declarator
+                name: (identifier) @name.definition.variable))) @definition.variable
+          
+          ;; === DESTRUCTURING DEFINITIONS ===
+          
+          ;; For-of/for-in loop with array destructuring: for (const [a, b] of items)
+          (for_in_statement
+            left: (array_pattern
+              (identifier) @name.definition.variable))
+          
+          ;; For-of/for-in loop with object destructuring: for (const { a, b } of items)
+          (for_in_statement
+            left: (object_pattern
+              (shorthand_property_identifier_pattern) @name.definition.variable))
+          
+          ;; For-of/for-in loop with object destructuring (aliased): for (const { a: b } of items)
+          (for_in_statement
+            left: (object_pattern
+              (pair_pattern
+                value: (identifier) @name.definition.variable)))
+          
+          ;; Regular array destructuring: const [a, b] = someArray
+          (lexical_declaration
+            (variable_declarator
+              name: (array_pattern
+                (identifier) @name.definition.variable)))
+          
+          ;; Regular object destructuring (shorthand): const { a, b } = someObject
+          (lexical_declaration
+            (variable_declarator
+              name: (object_pattern
+                (shorthand_property_identifier_pattern) @name.definition.variable)))
+          
+          ;; Regular object destructuring (aliased): const { a: b } = someObject
+          (lexical_declaration
+            (variable_declarator
+              name: (object_pattern
+                (pair_pattern
+                  value: (identifier) @name.definition.variable))))
+          
+          ;; Var array destructuring: var [a, b] = someArray
+          (variable_declaration
+            (variable_declarator
+              name: (array_pattern
+                (identifier) @name.definition.variable)))
+          
+          ;; Var object destructuring (shorthand): var { a, b } = someObject
+          (variable_declaration
+            (variable_declarator
+              name: (object_pattern
+                (shorthand_property_identifier_pattern) @name.definition.variable)))
+          
+          ;; Var object destructuring (aliased): var { a: b } = someObject
+          (variable_declaration
+            (variable_declarator
+              name: (object_pattern
+                (pair_pattern
+                  value: (identifier) @name.definition.variable))))
+          
+          ;; Function parameters with array destructuring: function foo([a, b]) {}
+          (formal_parameters
+            (required_parameter
+              pattern: (array_pattern
+                (identifier) @name.definition.variable)))
+          
+          ;; Function parameters with object destructuring (shorthand): function foo({ a, b }) {}
+          (formal_parameters
+            (required_parameter
+              pattern: (object_pattern
+                (shorthand_property_identifier_pattern) @name.definition.variable)))
+          
+          ;; Function parameters with object destructuring (aliased): function foo({ a: b }) {}
+          (formal_parameters
+            (required_parameter
+              pattern: (object_pattern
+                (pair_pattern
+                  value: (identifier) @name.definition.variable))))
+          
+          ;; Function parameters (simple): function foo(a, b) {}
+          (formal_parameters
+            (required_parameter
+              pattern: (identifier) @name.definition.variable))
+          
+          ;; === REFERENCES ===
+          
+          ;; Member expression object: foo.bar (capture foo as reference)
+          (member_expression
+            object: (identifier) @name.reference.variable)
+          
+          ;; Capture ALL type identifiers as references (comprehensive)
+          (type_identifier) @name.reference.type
+          
+          ;; new expressions: new SomeClass()
+          (new_expression
+            constructor: (identifier) @name.reference.class) @reference.class
+          
+          ;; Named imports (simple): import { DiffViewer } from './code-viewer'
+          (import_specifier
+            name: (identifier) @name.reference.variable
+            !alias) @reference.import
+          
+          ;; Named imports with alias: import { Foo as Bar } from './module'
+          (import_specifier
+            name: (identifier) @alias.original
+            alias: (identifier) @alias.name) @reference.import.alias
+          
+          ;; Default imports: import Foo from 'bar'
+          (import_clause
+            (identifier) @name.reference.import)
+          
+          ;; JSX opening element: <DiffViewer ...>
+          (jsx_opening_element
+            name: (identifier) @name.reference.class)
+          
+          ;; JSX self-closing element: <DiffViewer ... />
+          (jsx_self_closing_element
+            name: (identifier) @name.reference.class)
+          
+          ;; JSX member expression element: <Foo.Bar />
+          (jsx_opening_element
+            name: (member_expression
+              object: (identifier) @name.reference.variable))
+          
+          (jsx_self_closing_element
+            name: (member_expression
+              object: (identifier) @name.reference.variable))
+          
+          ;; Function calls: someFunction()
+          (call_expression
+            function: (identifier) @name.reference.call)
+          
+          ;; Method calls on objects: object.method()
+          (call_expression
+            function: (member_expression
+              object: (identifier) @name.reference.variable))
+          
+          ;; Extends clause in class: class Foo extends Bar
+          (class_heritage
+            (extends_clause
+              value: (identifier) @name.reference.class))
+          
+          ;; Implements clause: class Foo implements Bar
+          (class_heritage
+            (implements_clause
+              (type_identifier) @name.reference.interface))
+          
+          ;; Named exports: export { Foo }
+          (export_specifier
+            name: (identifier) @name.reference.export)
+          
+          ;; Array element identifiers: [foo, bar] (e.g., React useCallback/useEffect dependency arrays)
+          (array
+            (identifier) @name.reference.variable)
+        "#
+        .to_owned(),
+        || tree_sitter_typescript::LANGUAGE_TSX.into(),
+    )
+}

crates/codegen/kigi-codebase-graph/src/languages/types.rs

@@ -0,0 +1,82 @@
+use crate::scope_graph::nodes::SymbolId;
+
+/// Function type for getting a tree-sitter language grammar.
+pub type GrammarFn = fn() -> tree_sitter::Language;
+
+/// Contains information about the language and extra information which we would need
+/// per language
+pub struct TSLanguageConfig {
+    language_ids: Vec<String>,
+    file_extensions: Vec<String>,
+    namespaces: Vec<Vec<String>>,
+    file_definition_queries: String,
+    grammar: GrammarFn,
+}
+
+impl TSLanguageConfig {
+    pub fn new(
+        language_ids: Vec<String>,
+        file_extensions: Vec<String>,
+        namespaces: Vec<Vec<String>>,
+        file_definition_queries: String,
+        grammar: GrammarFn,
+    ) -> Self {
+        Self {
+            language_ids,
+            file_extensions,
+            namespaces,
+            file_definition_queries,
+            grammar,
+        }
+    }
+
+    /// Get the language IDs.
+    pub fn language_ids(&self) -> &[String] {
+        &self.language_ids
+    }
+
+    /// Get the first language ID, or "unknown" if none.
+    pub fn primary_language_id(&self) -> &str {
+        self.language_ids
+            .first()
+            .map(|s| s.as_str())
+            .unwrap_or("unknown")
+    }
+
+    /// Get the file extensions.
+    pub fn file_extensions(&self) -> &[String] {
+        &self.file_extensions
+    }
+
+    /// Get the namespaces.
+    pub fn namespaces(&self) -> &[Vec<String>] {
+        &self.namespaces
+    }
+
+    /// Get the file definition queries.
+    pub fn file_definition_queries(&self) -> &str {
+        &self.file_definition_queries
+    }
+
+    /// Get the tree-sitter language.
+    pub fn language(&self) -> tree_sitter::Language {
+        (self.grammar)()
+    }
+
+    /// Compile the definitions query for this language.
+    pub fn compile_query(&self) -> Result<tree_sitter::Query, tree_sitter::QueryError> {
+        tree_sitter::Query::new(&self.language(), &self.file_definition_queries)
+    }
+
+    /// Find a SymbolId for a given symbol type name.
+    pub fn symbol_id_of(&self, symbol_type: &str) -> Option<SymbolId> {
+        for (ns_idx, namespace) in self.namespaces.iter().enumerate() {
+            for (sym_idx, sym) in namespace.iter().enumerate() {
+                if sym == symbol_type {
+                    return Some(SymbolId::new(ns_idx, sym_idx));
+                }
+            }
+        }
+        None
+    }
+}

crates/codegen/kigi-codebase-graph/src/lib.rs

@@ -0,0 +1,102 @@
+//! # kigi-codebase-graph
+//!
+//! High-performance code graph generation using tree-sitter queries.
+//!
+//! This crate provides:
+//! - **Go-to-definitions**: Find where symbols are defined
+//! - **Go-to-references**: Find where symbols are used
+//! - **Initial repository indexing**: Build the full index from scratch
+//! - **Incremental reindexing**: Update the index based on file system events
+//! - **Parallel processing**: Uses rayon for fast parallel parsing
+//! - **Memory-mapped I/O**: Zero-copy file reading and fast index caching
+//!
+//! ## Quick Start
+//!
+//! ```rust,ignore
+//! use std::path::Path;
+//! use kigi_codebase_graph::{IndexBuilder, load_index, save_index, get_cache_path, Navigator};
+//!
+//! let repo_path = Path::new("/path/to/repo");
+//! let cache_path = get_cache_path(repo_path);
+//!
+//! // Try loading from cache first, otherwise build fresh
+//! let index = match load_index(&cache_path) {
+//!     Ok(index) => index,
+//!     Err(_) => {
+//!         let index = IndexBuilder::new()
+//!             .with_threads(8)
+//!             .build(repo_path)?;
+//!         save_index(&cache_path, &index)?;
+//!         index
+//!     }
+//! };
+//!
+//! // Create a navigator for location-based operations
+//! let navigator = Navigator::new(index);
+//!
+//! // Go to definition at a specific position (row and col are 1-indexed)
+//! let result = navigator.goto_definition(Path::new("src/main.rs"), 10, 15)?;
+//! for loc in result.locations {
+//!     println!("{}:{}", loc.path.display(), loc.line);
+//! }
+//! ```
+//!
+//! ## Channel-Based Incremental Updates
+//!
+//! `IndexManagerHandle` exposes direct query commands that answer in-place
+//! without cloning the full index.  Prefer these over `get_snapshot()` in
+//! hot paths.
+//!
+//! ```rust,ignore
+//! use std::path::PathBuf;
+//! use kigi_codebase_graph::{IndexManager, IndexManagerConfig, FileEvent};
+//!
+//! // Create the manager with config
+//! let config = IndexManagerConfig::new("/path/to/repo".into())
+//!     .with_cache_path("/tmp/index.bin".into());
+//!
+//! let handle = IndexManager::spawn(config);
+//!
+//! // Send file events as they come from FSNotify
+//! handle.send_event(FileEvent::modified("src/main.rs".into()))?;
+//!
+//! // Query directly — no full-index clone needed
+//! let file = PathBuf::from("src/main.rs");
+//! let result = handle.goto_definition_blocking(file, 10, 15)??;
+//! for loc in result.locations {
+//!     println!("{}:{}", loc.path, loc.line);
+//! }
+//!
+//! // Lightweight stats — also no clone
+//! let file_count = handle.get_file_count();
+//! let exists     = handle.has_definition_blocking("MyStruct");
+//! ```
+
+pub mod index_manager;
+pub mod interner;
+pub mod languages;
+pub mod manager;
+pub mod navigation;
+pub mod scope_graph;
+pub mod types;
+
+// Re-exports for convenient access
+pub use index_manager::{
+    FileEvent, FileEventKind, IndexCommand, IndexManager, IndexManagerConfig, IndexManagerHandle,
+    MAX_INDEXABLE_FILE_SIZE, QueryError, QueryResult, SymbolLocation, is_binary_content,
+};
+pub use languages::{LanguageRegistry, TSLanguageConfig};
+pub use manager::{
+    CACHE_FILE_NAME, CacheError, IndexBuilder, IndexError, IndexOperation, LockResult,
+    WorkspaceLockGuard, cache_exists, cache_size, get_cache_path, is_operation_in_progress,
+    load_index, save_index, save_index_async, try_lock,
+};
+pub use navigation::{Location, NavigationError, NavigationResult, Navigator};
+pub use scope_graph::{
+    LocalDef, LocalImport, LocalScope, NodeKind, QueryVersion, Reference, ScopeGraph,
+    ScopeGraphIndex, ScopeGraphResult, Symbol, SymbolId, build_scope_graph, extract_symbols_fast,
+};
+pub use types::{FileMeta, IndexStats, Position, Range, SymbolAlias, SymbolOccurrence};
+
+// String interning for memory-efficient storage
+pub use interner::{StringId, StringInterner};

crates/codegen/kigi-codebase-graph/src/manager/builder.rs

@@ -0,0 +1,507 @@
+//! Parallel pipelined index builder with thread-local caching.
+
+use std::cell::RefCell;
+use std::fs;
+use std::path::{Path, PathBuf};
+use std::sync::Arc;
+
+use ahash::AHashMap as HashMap;
+use ignore::WalkBuilder;
+use rayon::prelude::*;
+
+use crate::languages::LanguageRegistry;
+use crate::scope_graph::ScopeGraphIndex;
+use crate::types::{FileMeta, SymbolAlias, SymbolOccurrence};
+use kigi_paths::to_relative_path;
+
+/// Error type for index building operations.
+#[derive(Debug)]
+pub enum IndexError {
+    /// Error walking directory.
+    WalkError { message: String },
+    /// Thread panicked.
+    ThreadPanic { message: String },
+    /// IO error.
+    IoError(std::io::Error),
+}
+
+impl std::fmt::Display for IndexError {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            IndexError::WalkError { message } => write!(f, "Walk error: {}", message),
+            IndexError::ThreadPanic { message } => write!(f, "Thread panic: {}", message),
+            IndexError::IoError(e) => write!(f, "IO error: {}", e),
+        }
+    }
+}
+
+impl std::error::Error for IndexError {}
+
+impl From<std::io::Error> for IndexError {
+    fn from(e: std::io::Error) -> Self {
+        IndexError::IoError(e)
+    }
+}
+
+/// Result type for index building operations.
+pub type Result<T> = std::result::Result<T, IndexError>;
+
+// Thread-local caches for parsers and queries to avoid repeated initialization
+thread_local! {
+    static PARSER_CACHE: RefCell<HashMap<String, tree_sitter::Parser>> = RefCell::new(HashMap::new());
+    static QUERY_CACHE: RefCell<HashMap<String, tree_sitter::Query>> = RefCell::new(HashMap::new());
+}
+
+/// Extracted symbols for a single file (lightweight, no ScopeGraph overhead)
+struct FileSymbols {
+    path: Arc<str>,
+    definitions: Vec<SymbolOccurrence>,
+    references: Vec<SymbolOccurrence>,
+    aliases: Vec<SymbolAlias>,
+    file_meta: FileMeta,
+}
+
+/// Builder for creating a symbol index.
+///
+/// Uses optimized parallel processing with:
+/// - Thread-local parser and query caching
+/// - Chunked parallel processing for cache locality
+/// - Lightweight symbol extraction (no intermediate ScopeGraph)
+/// - Bounded merge-batching to cap two-phase peak memory
+pub struct IndexBuilder {
+    registry: LanguageRegistry,
+    num_threads: usize,
+    /// Whether to respect .gitignore files (default: true)
+    respect_gitignore: bool,
+    /// Whether to skip hidden files/directories (default: true)
+    skip_hidden: bool,
+    /// Chunk size for parallel processing / thread-local cache locality (default: 100)
+    chunk_size: usize,
+    /// Maximum number of files whose symbols are held in memory at once during
+    /// the merge phase.  Limiting this bounds the two-phase peak: parallel
+    /// parsing produces at most `build_batch_size` FileSymbols before they are
+    /// merged into the index and dropped.  Default: 5 000 files per batch.
+    build_batch_size: usize,
+}
+
+/// Get the default number of threads (N-1 cores, minimum 1).
+fn default_num_threads() -> usize {
+    num_cpus::get().saturating_sub(1).max(1)
+}
+
+impl IndexBuilder {
+    /// Create a new index builder.
+    pub fn new() -> Self {
+        Self {
+            registry: LanguageRegistry::new(),
+            num_threads: default_num_threads(),
+            respect_gitignore: true,
+            skip_hidden: true,
+            chunk_size: 100,
+            build_batch_size: 5_000,
+        }
+    }
+
+    /// Create a new index builder with a custom language registry.
+    pub fn with_registry(registry: LanguageRegistry) -> Self {
+        Self {
+            registry,
+            num_threads: default_num_threads(),
+            respect_gitignore: true,
+            skip_hidden: true,
+            chunk_size: 100,
+            build_batch_size: 5_000,
+        }
+    }
+
+    /// Set the number of threads to use (default: N-1 cores).
+    #[must_use]
+    pub fn with_threads(mut self, count: usize) -> Self {
+        self.num_threads = count;
+        self
+    }
+
+    /// Set the chunk size for parallel processing (default: 100).
+    #[must_use]
+    pub fn with_chunk_size(mut self, size: usize) -> Self {
+        self.chunk_size = size;
+        self
+    }
+
+    /// Set the merge-batch size (default: 5 000 files per batch).
+    ///
+    /// Controls how many files' symbols are held in memory simultaneously
+    /// during the sequential merge phase.  Smaller values reduce peak RSS at
+    /// the cost of slightly more pool scheduling overhead.  Values below
+    /// `chunk_size` are clamped to `chunk_size` at build time, so the call
+    /// order of `with_build_batch_size` and `with_chunk_size` does not matter.
+    #[must_use]
+    pub fn with_build_batch_size(mut self, size: usize) -> Self {
+        self.build_batch_size = size;
+        self
+    }
+
+    /// Set whether to respect .gitignore files (default: true).
+    #[must_use]
+    pub fn respect_gitignore(mut self, respect: bool) -> Self {
+        self.respect_gitignore = respect;
+        self
+    }
+
+    /// Set whether to skip hidden files/directories (default: true).
+    #[must_use]
+    pub fn skip_hidden(mut self, skip: bool) -> Self {
+        self.skip_hidden = skip;
+        self
+    }
+
+    /// Build index from a directory, respecting .gitignore.
+    ///
+    /// This walks the directory tree, automatically respecting:
+    /// - `.gitignore` files at any level
+    /// - `.git/info/exclude`
+    /// - Global gitignore (`~/.config/git/ignore`)
+    /// - Hidden files/directories (configurable)
+    ///
+    /// **Note**: File paths in the index are stored as **relative paths** (to `root_path`)
+    /// for portability across machines/sessions.
+    pub fn build(&self, root_path: &Path) -> Result<ScopeGraphIndex> {
+        // Collect files using the ignore crate
+        let file_paths = self.collect_files(root_path)?;
+
+        if file_paths.is_empty() {
+            let mut index = ScopeGraphIndex::new();
+            // Set query version even for empty index so cache validation works
+            index.set_query_version(self.registry.compute_query_hash());
+            return Ok(index);
+        }
+
+        self.build_fast(root_path, &file_paths)
+    }
+
+    /// Collect all supported files from a directory, respecting gitignore.
+    /// Uses `git ls-files` when available (faster), falls back to directory walking.
+    fn collect_files(&self, root_path: &Path) -> Result<Vec<PathBuf>> {
+        // Try git ls-files first - it's much faster as it reads from git's index
+        // But it only works for tracked files, so we also add untracked files
+        if let Some(files) = self.collect_files_git(root_path)
+            && !files.is_empty()
+        {
+            return Ok(files);
+        }
+
+        // Fall back to directory walking
+        self.collect_files_walk(root_path)
+    }
+
+    /// Collect files using git2 - reads from the git index (tracked files).
+    /// Untracked files are not included since:
+    /// 1. They are typically a small minority
+    /// 2. They will be picked up by fsnotify when created
+    /// 3. The statuses() call for untracked files is very slow (~10x overhead)
+    fn collect_files_git(&self, root_path: &Path) -> Option<Vec<PathBuf>> {
+        use git2::Repository;
+
+        // Open the repository
+        let repo = Repository::open(root_path).ok()?;
+
+        // Get all files from the index (tracked files)
+        let index = repo.index().ok()?;
+        let files: Vec<PathBuf> = index
+            .iter()
+            .filter_map(|entry| {
+                // git2 stores paths as bytes, convert to str
+                let path_str = std::str::from_utf8(&entry.path).ok()?;
+                if self.registry.is_supported(path_str) {
+                    Some(root_path.join(path_str))
+                } else {
+                    None
+                }
+            })
+            .collect();
+
+        Some(files)
+    }
+
+    /// Collect files by walking the directory tree.
+    /// Used as fallback when not in a git repository.
+    fn collect_files_walk(&self, root_path: &Path) -> Result<Vec<PathBuf>> {
+        use std::sync::Mutex;
+
+        let files = Mutex::new(Vec::with_capacity(50000));
+
+        let walker = WalkBuilder::new(root_path)
+            .hidden(self.skip_hidden)
+            .git_ignore(self.respect_gitignore)
+            .git_global(self.respect_gitignore)
+            .git_exclude(self.respect_gitignore)
+            .threads(self.num_threads.min(12)) // Use parallel walking (capped at 12)
+            .build_parallel();
+
+        walker.run(|| {
+            let files = &files;
+            let registry = &self.registry;
+            Box::new(move |entry| {
+                use ignore::WalkState;
+
+                let entry = match entry {
+                    Ok(e) => e,
+                    Err(_) => return WalkState::Continue,
+                };
+
+                let path = entry.path();
+
+                // Skip directories
+                if path.is_dir() {
+                    return WalkState::Continue;
+                }
+
+                // Check if the file is supported
+                if registry.is_supported(path) {
+                    files.lock().unwrap().push(path.to_path_buf());
+                }
+
+                WalkState::Continue
+            })
+        });
+
+        Ok(files.into_inner().unwrap())
+    }
+
+    /// Build index with maximum throughput optimizations:
+    /// - Memory-mapped I/O for zero-copy file reading
+    /// - Direct parsing from mmap (no intermediate buffer copy)
+    /// - Lightweight symbol extraction (skip building full ScopeGraph)
+    /// - Thread-local parser and query caching
+    /// - Chunked parallel processing for better cache locality
+    /// - Single StringInterner for memory-efficient string deduplication
+    ///
+    /// Uses two-phase approach:
+    /// 1. Parallel: parse files and extract symbols into Vec<FileSymbols>
+    /// 2. Sequential: aggregate into single ScopeGraphIndex with single interner
+    ///
+    /// This ensures all strings are deduplicated in one interner, avoiding
+    /// the memory overhead of multiple interners during parallel aggregation.
+    ///
+    /// File paths are stored as **relative paths** (to `root_path`) for portability.
+    fn build_fast(&self, root_path: &Path, file_paths: &[PathBuf]) -> Result<ScopeGraphIndex> {
+        // Configure thread pool with N-1 cores
+        let pool = rayon::ThreadPoolBuilder::new()
+            .num_threads(self.num_threads)
+            .build()
+            .map_err(|e| IndexError::WalkError {
+                message: format!("Failed to create thread pool: {}", e),
+            })?;
+
+        let registry = Arc::new(LanguageRegistry::new());
+        let chunk_size = self.chunk_size;
+        // Clamp here against the final chunk_size so that call order of
+        // with_build_batch_size / with_chunk_size on the builder does not matter.
+        let build_batch_size = self.build_batch_size.max(chunk_size);
+        let root_arc: Arc<Path> = Arc::from(root_path);
+
+        let mut index = ScopeGraphIndex::new();
+
+        // Process files in bounded merge-batches to cap two-phase peak memory.
+        //
+        // Old approach: collect ALL FileSymbols in one go, then merge.
+        // Peak = O(total_files) symbols + growing index simultaneously.
+        //
+        // New approach: for each batch of build_batch_size files:
+        //   1. Parse in parallel (par_chunks preserves thread-local cache locality)
+        //   2. Merge the batch into the index
+        //   3. Drop the batch before starting the next one
+        // Peak = O(build_batch_size) symbols + growing index simultaneously.
+        for batch in file_paths.chunks(build_batch_size) {
+            let batch_symbols: Vec<FileSymbols> = pool.install(|| {
+                batch
+                    .par_chunks(chunk_size)
+                    .flat_map_iter(|chunk| {
+                        chunk
+                            .iter()
+                            .filter_map(|path| process_file_fast(path, &root_arc, &registry))
+                    })
+                    .collect()
+            });
+
+            for file_syms in batch_symbols {
+                let path_str: &str = &file_syms.path;
+                for sym in file_syms.definitions {
+                    index.add_definition(&sym.name, path_str, sym.line);
+                }
+                for sym in file_syms.references {
+                    index.add_reference(&sym.name, path_str, sym.line);
+                }
+                for alias in file_syms.aliases {
+                    index.add_alias_arc(alias.alias, alias.original);
+                }
+                index.set_file_meta(path_str, file_syms.file_meta);
+            }
+            // batch_symbols dropped here — frees the parallel-extracted symbols
+            // before the next batch is parsed
+        }
+
+        // Set the query version hash so we can detect query changes on cache load
+        index.set_query_version(self.registry.compute_query_hash());
+
+        // Reclaim over-allocated Vec capacity that accumulated during bulk push().
+        // This is a one-time cost paid here (O(symbols)) to permanently reduce RSS.
+        index.compact();
+
+        Ok(index)
+    }
+}
+
+impl Default for IndexBuilder {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+/// Process a single file using thread-local caching.
+/// Returns lightweight FileSymbols (no ScopeGraph overhead).
+///
+/// File path is stored as **relative** (to `root_path`) for portability.
+fn process_file_fast(
+    path: &Path,
+    root_path: &Path,
+    registry: &LanguageRegistry,
+) -> Option<FileSymbols> {
+    use crate::index_manager::MAX_INDEXABLE_FILE_SIZE;
+
+    let lang_config = registry.for_file_path(path)?;
+    let lang_id = lang_config.primary_language_id().to_string();
+
+    let metadata = fs::metadata(path).ok()?;
+    if metadata.len() == 0 || metadata.len() > MAX_INDEXABLE_FILE_SIZE {
+        return None;
+    }
+
+    // Prefix-read binary check: only reads 8KB, not the whole file
+    {
+        use std::io::Read;
+        let mut f = fs::File::open(path).ok()?;
+        let mut buf = [0u8; 8000];
+        let n = f.read(&mut buf).ok()?;
+        if buf[..n].contains(&0) {
+            return None;
+        }
+    }
+
+    let content = fs::read(path).ok()?;
+
+    // Parse using thread-local cached parser
+    let tree = PARSER_CACHE.with(|cache| {
+        let mut cache = cache.borrow_mut();
+        let parser = cache.entry(lang_id.clone()).or_insert_with(|| {
+            let mut p = tree_sitter::Parser::new();
+            let ts_lang = lang_config.language();
+            let _ = p.set_language(&ts_lang);
+            p
+        });
+        parser.parse(&content, None)
+    })?;
+
+    let root_node = tree.root_node();
+
+    // Extract symbols using thread-local cached query
+    let (definitions, references, aliases) = QUERY_CACHE.with(|cache| {
+        let mut cache = cache.borrow_mut();
+        let query = cache.entry(lang_id.clone()).or_insert_with(|| {
+            lang_config.compile_query().unwrap_or_else(|_| {
+                let ts_lang = lang_config.language();
+                tree_sitter::Query::new(&ts_lang, "").expect("empty query should always work")
+            })
+        });
+        extract_symbols_fast_inline(query, root_node, &content)
+    });
+
+    // Reuse metadata from the size check above (no re-stat needed)
+
+    // Convert absolute path to relative for portable storage
+    let rel_path = to_relative_path(root_path, path);
+
+    Some(FileSymbols {
+        path: rel_path.to_string_lossy().into(),
+        definitions,
+        references,
+        aliases,
+        file_meta: FileMeta::from_metadata(&metadata),
+    })
+}
+
+/// Lightweight symbol extraction - returns proper typed vectors.
+/// Inlined for maximum performance (avoids function call overhead in hot loop).
+#[inline]
+fn extract_symbols_fast_inline(
+    query: &tree_sitter::Query,
+    root_node: tree_sitter::Node<'_>,
+    src: &[u8],
+) -> (
+    Vec<SymbolOccurrence>,
+    Vec<SymbolOccurrence>,
+    Vec<SymbolAlias>,
+) {
+    use tree_sitter::StreamingIterator;
+
+    // Pre-compute capture indices for fast lookup
+    let capture_names = query.capture_names();
+    let mut is_def = vec![false; capture_names.len()];
+    let mut is_ref = vec![false; capture_names.len()];
+    let mut alias_original_idx: Option<usize> = None;
+    let mut alias_name_idx: Option<usize> = None;
+
+    for (i, name) in capture_names.iter().enumerate() {
+        if name.starts_with("name.definition.") {
+            is_def[i] = true;
+        } else if name.starts_with("name.reference.") {
+            is_ref[i] = true;
+        } else if *name == "alias.original" {
+            alias_original_idx = Some(i);
+        } else if *name == "alias.name" {
+            alias_name_idx = Some(i);
+        }
+    }
+
+    // Pre-allocate with reasonable capacity
+    let mut definitions: Vec<SymbolOccurrence> = Vec::with_capacity(64);
+    let mut references: Vec<SymbolOccurrence> = Vec::with_capacity(256);
+    let mut aliases: Vec<SymbolAlias> = Vec::with_capacity(8);
+
+    let mut cursor = tree_sitter::QueryCursor::new();
+    let mut matches = cursor.matches(query, root_node, src);
+
+    while let Some(match_) = matches.next() {
+        let mut alias_original: Option<&[u8]> = None;
+        let mut alias_name: Option<&[u8]> = None;
+
+        for capture in match_.captures {
+            let idx = capture.index as usize;
+            let node = capture.node;
+            let byte_range = node.byte_range();
+
+            if is_def.get(idx).copied().unwrap_or(false) {
+                // Convert Cow<str> directly to Arc<str> - avoids intermediate String allocation
+                let text: Arc<str> = String::from_utf8_lossy(&src[byte_range]).into();
+                // Line numbers are 1-indexed
+                definitions.push(SymbolOccurrence::new(text, node.start_position().row + 1));
+            } else if is_ref.get(idx).copied().unwrap_or(false) {
+                let text: Arc<str> = String::from_utf8_lossy(&src[byte_range]).into();
+                references.push(SymbolOccurrence::new(text, node.start_position().row + 1));
+            } else if Some(idx) == alias_original_idx {
+                alias_original = Some(&src[byte_range]);
+            } else if Some(idx) == alias_name_idx {
+                alias_name = Some(&src[byte_range]);
+            }
+        }
+
+        if let (Some(original), Some(alias)) = (alias_original, alias_name) {
+            // Convert Cow<str> directly to Arc<str> - avoids intermediate String allocation
+            let orig_arc: Arc<str> = String::from_utf8_lossy(original).into();
+            let alias_arc: Arc<str> = String::from_utf8_lossy(alias).into();
+            aliases.push(SymbolAlias::new(alias_arc, orig_arc));
+        }
+    }
+
+    (definitions, references, aliases)
+}

crates/codegen/kigi-codebase-graph/src/manager/cache.rs

@@ -0,0 +1,106 @@
+//! Index caching for fast loading.
+//!
+//! Uses a custom binary format with magic bytes "SGIX" for the new interned format.
+//! Automatically detects and skips legacy bincode format (returns error so caller can rebuild).
+
+use std::path::Path;
+
+use crate::scope_graph::ScopeGraphIndex;
+
+/// Default cache file name.
+pub const CACHE_FILE_NAME: &str = ".goto_index.bin";
+
+/// Error type for cache operations.
+#[derive(Debug)]
+pub enum CacheError {
+    /// IO error.
+    IoError(std::io::Error),
+    /// Serialization error.
+    SerializeError(String),
+    /// Deserialization error.
+    DeserializeError(String),
+    /// Legacy format detected (caller should rebuild).
+    LegacyFormat,
+}
+
+impl std::fmt::Display for CacheError {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            CacheError::IoError(e) => write!(f, "IO error: {}", e),
+            CacheError::SerializeError(msg) => write!(f, "Serialization error: {}", msg),
+            CacheError::DeserializeError(msg) => write!(f, "Deserialization error: {}", msg),
+            CacheError::LegacyFormat => write!(f, "Legacy cache format detected"),
+        }
+    }
+}
+
+impl std::error::Error for CacheError {}
+
+impl From<std::io::Error> for CacheError {
+    fn from(e: std::io::Error) -> Self {
+        CacheError::IoError(e)
+    }
+}
+
+/// Result type for cache operations.
+pub type Result<T> = std::result::Result<T, CacheError>;
+
+/// Get the default cache path for a repository.
+pub fn get_cache_path(root_path: &Path) -> std::path::PathBuf {
+    root_path.join(CACHE_FILE_NAME)
+}
+
+/// Load an index from cache.
+///
+/// Uses the new binary format with magic bytes "SGIX".
+/// Returns `CacheError::LegacyFormat` if the file uses the old bincode format,
+/// signaling to the caller that a rebuild is needed.
+pub fn load_index(cache_path: &Path) -> Result<ScopeGraphIndex> {
+    if !cache_path.exists() {
+        return Err(CacheError::IoError(std::io::Error::new(
+            std::io::ErrorKind::NotFound,
+            "Cache file not found",
+        )));
+    }
+
+    // Use ScopeGraphIndex::load which handles format detection
+    match ScopeGraphIndex::load(cache_path) {
+        Ok(Some(index)) => Ok(index),
+        Ok(None) => {
+            // None means legacy format was detected
+            tracing::info!(
+                cache_path = %cache_path.display(),
+                "Legacy cache format detected, will rebuild"
+            );
+            Err(CacheError::LegacyFormat)
+        }
+        Err(e) => Err(CacheError::IoError(e)),
+    }
+}
+
+/// Save an index to cache using the new binary format.
+pub fn save_index(cache_path: &Path, index: &ScopeGraphIndex) -> Result<()> {
+    index.save(cache_path).map_err(CacheError::IoError)
+}
+
+/// Save an index to cache asynchronously (in a background thread).
+///
+/// Returns immediately and spawns a thread to do the actual saving.
+/// Useful for saving the index without blocking the main thread.
+pub fn save_index_async(cache_path: std::path::PathBuf, index: ScopeGraphIndex) {
+    std::thread::spawn(move || {
+        if let Err(e) = save_index(&cache_path, &index) {
+            tracing::warn!("Failed to save index cache: {}", e);
+        }
+    });
+}
+
+/// Check if a cache exists and return its metadata.
+pub fn cache_exists(cache_path: &Path) -> bool {
+    cache_path.exists()
+}
+
+/// Get cache file size in bytes.
+pub fn cache_size(cache_path: &Path) -> Option<u64> {
+    std::fs::metadata(cache_path).ok().map(|m| m.len())
+}

crates/codegen/kigi-codebase-graph/src/manager/lock.rs

@@ -0,0 +1,539 @@
+//! Workspace-level locking for index operations.
+//!
+//! Provides both in-memory (same-process) and file-based (cross-process)
+//! coordination to prevent redundant index operations on the same workspace.
+//!
+//! ## Design
+//!
+//! - **In-memory locks**: Fast path for same-process deduplication using a global registry
+//! - **File locks**: Cross-process coordination using lock files with PID and timestamp
+//! - **Stale detection**: Locks are considered stale if the holding process is dead or timeout exceeded
+//!
+//! ## Lock Types
+//!
+//! - **Shared (Load)**: Multiple readers allowed, blocked during exclusive operations
+//! - **Exclusive (Save/Build/Refresh)**: Single writer, blocks all other operations
+
+use std::path::{Path, PathBuf};
+use std::time::{Duration, SystemTime, UNIX_EPOCH};
+
+use dashmap::DashMap;
+use once_cell::sync::Lazy;
+
+const LOAD_STALE_DURATION_SEC: u64 = 120;
+const SAVE_STALE_DURATION_SEC: u64 = 120;
+const BUILD_STALE_DURATION_SEC: u64 = 600;
+const BG_REFRESH_STALE_DURATION_SEC: u64 = 300;
+
+/// Operations that require locking.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum IndexOperation {
+    /// Loading index from cache (shared/read lock).
+    Load,
+    /// Saving index to cache (exclusive).
+    Save,
+    /// Building index from scratch (exclusive).
+    Build,
+    /// Background validation and refresh (exclusive).
+    BackgroundRefresh,
+}
+
+impl IndexOperation {
+    /// String representation for lock file.
+    fn as_str(&self) -> &'static str {
+        match self {
+            Self::Load => "load",
+            Self::Save => "save",
+            Self::Build => "build",
+            Self::BackgroundRefresh => "background_refresh",
+        }
+    }
+
+    /// Whether this operation requires exclusive access.
+    pub fn is_exclusive(&self) -> bool {
+        match self {
+            Self::Load => false, // Shared/read access
+            Self::Save | Self::Build | Self::BackgroundRefresh => true,
+        }
+    }
+
+    /// Timeout after which a lock is considered stale.
+    fn stale_timeout(&self) -> Duration {
+        match self {
+            Self::Load => Duration::from_secs(LOAD_STALE_DURATION_SEC),
+            Self::Save => Duration::from_secs(SAVE_STALE_DURATION_SEC),
+            Self::Build => Duration::from_secs(BUILD_STALE_DURATION_SEC),
+            Self::BackgroundRefresh => Duration::from_secs(BG_REFRESH_STALE_DURATION_SEC),
+        }
+    }
+}
+
+impl std::fmt::Display for IndexOperation {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "{}", self.as_str())
+    }
+}
+
+/// In-memory lock state for same-process deduplication.
+struct InMemoryLockState {
+    operation: IndexOperation,
+    readers: usize,  // Count for shared locks
+    exclusive: bool, // Whether an exclusive lock is held
+}
+
+/// Global registry of in-memory locks (same process).
+/// Uses DashMap for lock-free concurrent access.
+static IN_MEMORY_LOCKS: Lazy<DashMap<PathBuf, InMemoryLockState>> = Lazy::new(DashMap::new);
+
+/// A guard that releases the lock when dropped.
+pub struct WorkspaceLockGuard {
+    workspace: PathBuf,
+    lock_file_path: PathBuf,
+    operation: IndexOperation,
+}
+
+impl Drop for WorkspaceLockGuard {
+    fn drop(&mut self) {
+        // Release in-memory lock
+        release_in_memory_lock(&self.workspace, self.operation);
+
+        // Release file lock (only for exclusive operations)
+        if self.operation.is_exclusive()
+            && let Err(e) = std::fs::remove_file(&self.lock_file_path)
+            && e.kind() != std::io::ErrorKind::NotFound
+        {
+            tracing::warn!(
+                path = %self.lock_file_path.display(),
+                error = %e,
+                "Failed to remove lock file"
+            );
+        }
+    }
+}
+
+/// Result of trying to acquire a lock.
+pub enum LockResult {
+    /// Lock acquired successfully.
+    Acquired(WorkspaceLockGuard),
+    /// Another operation is in progress.
+    Busy {
+        /// Description of the blocking operation.
+        operation: String,
+        /// PID of the process holding the lock (if known).
+        holder_pid: Option<u32>,
+    },
+}
+
+impl LockResult {
+    /// Returns true if the lock was acquired.
+    pub fn is_acquired(&self) -> bool {
+        matches!(self, Self::Acquired(_))
+    }
+
+    /// Unwrap the guard, panicking if busy.
+    pub fn unwrap(self) -> WorkspaceLockGuard {
+        match self {
+            Self::Acquired(guard) => guard,
+            Self::Busy { operation, .. } => {
+                panic!("Lock was busy: {}", operation)
+            }
+        }
+    }
+}
+
+/// Try to acquire a lock for an index operation on a workspace.
+///
+/// Returns `LockResult::Acquired` with a guard if successful, or `LockResult::Busy`
+/// if another operation is in progress.
+///
+/// # Arguments
+///
+/// * `workspace` - The workspace root path
+/// * `operation` - The type of operation to perform
+///
+/// # Example
+///
+/// ```ignore
+/// use kigi_codebase_graph::manager::lock::{try_lock, IndexOperation, LockResult};
+///
+/// let workspace = Path::new("/path/to/workspace");
+/// match try_lock(workspace, IndexOperation::Build) {
+///     LockResult::Acquired(guard) => {
+///         // Do work...
+///         // Lock is released when guard is dropped
+///     }
+///     LockResult::Busy { operation, holder_pid } => {
+///         println!("Busy: {} by pid {:?}", operation, holder_pid);
+///     }
+/// }
+/// ```
+pub fn try_lock(workspace: &Path, operation: IndexOperation) -> LockResult {
+    let workspace = canonicalize_workspace(workspace);
+    let lock_file_path = get_lock_file_path(&workspace);
+
+    // Step 1: Check/acquire in-memory lock (fast path for same process)
+    if !try_acquire_in_memory_lock(&workspace, operation) {
+        tracing::debug!(
+            workspace = %workspace.display(),
+            operation = %operation,
+            "In-memory lock busy"
+        );
+        return LockResult::Busy {
+            operation: format!("{} (same process)", operation),
+            holder_pid: Some(std::process::id()),
+        };
+    }
+
+    // Step 2: For exclusive operations, also acquire file lock (cross-process)
+    if operation.is_exclusive() {
+        match try_acquire_file_lock(&lock_file_path, operation) {
+            Ok(()) => {
+                tracing::debug!(
+                    workspace = %workspace.display(),
+                    operation = %operation,
+                    lock_file = %lock_file_path.display(),
+                    "Acquired exclusive lock"
+                );
+            }
+            Err((op, pid)) => {
+                // Release in-memory lock since we failed to get file lock
+                release_in_memory_lock(&workspace, operation);
+                tracing::debug!(
+                    workspace = %workspace.display(),
+                    operation = %operation,
+                    blocking_op = %op,
+                    blocking_pid = ?pid,
+                    "File lock busy"
+                );
+                return LockResult::Busy {
+                    operation: op,
+                    holder_pid: pid,
+                };
+            }
+        }
+    }
+
+    LockResult::Acquired(WorkspaceLockGuard {
+        workspace,
+        lock_file_path,
+        operation,
+    })
+}
+
+/// Check if an operation is currently in progress for a workspace.
+///
+/// This is a non-blocking check that doesn't acquire any locks.
+pub fn is_operation_in_progress(workspace: &Path, operation: IndexOperation) -> bool {
+    let workspace = canonicalize_workspace(workspace);
+
+    // Check in-memory first
+    if let Some(state) = IN_MEMORY_LOCKS.get(&workspace) {
+        if operation.is_exclusive() {
+            if state.readers > 0 || state.exclusive {
+                return true;
+            }
+        } else if state.exclusive {
+            return true;
+        }
+    }
+
+    // Check file lock for exclusive operations
+    if operation.is_exclusive()
+        && let Ok(contents) = std::fs::read_to_string(get_lock_file_path(&workspace))
+        && let Some((_, pid, started)) = parse_lock_file(&contents)
+    {
+        let age = SystemTime::now()
+            .duration_since(started)
+            .unwrap_or(Duration::ZERO);
+        if age < operation.stale_timeout() && is_process_alive(pid) {
+            return true;
+        }
+    }
+
+    false
+}
+
+/// Canonicalize workspace path for consistent lock keys.
+fn canonicalize_workspace(workspace: &Path) -> PathBuf {
+    // Try to canonicalize, fall back to the original path
+    dunce::canonicalize(workspace).unwrap_or_else(|_| workspace.to_path_buf())
+}
+
+/// Get the lock file path for a workspace.
+fn get_lock_file_path(workspace: &Path) -> PathBuf {
+    // Use the cache directory (same as where .goto_index.bin is stored)
+    let cache_path = super::get_cache_path(workspace);
+    cache_path.with_extension("lock")
+}
+
+/// Try to acquire an in-memory lock for same-process deduplication.
+fn try_acquire_in_memory_lock(workspace: &Path, operation: IndexOperation) -> bool {
+    // Use entry API for atomic check-and-modify
+    match IN_MEMORY_LOCKS.entry(workspace.to_path_buf()) {
+        dashmap::mapref::entry::Entry::Occupied(mut entry) => {
+            let state = entry.get_mut();
+            if operation.is_exclusive() {
+                // Exclusive operation - must have no readers or existing exclusive
+                if state.readers > 0 || state.exclusive {
+                    return false;
+                }
+                state.exclusive = true;
+                state.operation = operation;
+            } else {
+                // Shared operation (Load) - OK if no exclusive lock
+                if state.exclusive {
+                    return false;
+                }
+                state.readers += 1;
+            }
+        }
+        dashmap::mapref::entry::Entry::Vacant(entry) => {
+            // No existing lock - create one
+            entry.insert(InMemoryLockState {
+                operation,
+                readers: if operation.is_exclusive() { 0 } else { 1 },
+                exclusive: operation.is_exclusive(),
+            });
+        }
+    }
+    true
+}
+
+/// Release an in-memory lock.
+fn release_in_memory_lock(workspace: &Path, operation: IndexOperation) {
+    // Use entry API for atomic check-and-modify
+    if let dashmap::mapref::entry::Entry::Occupied(mut entry) =
+        IN_MEMORY_LOCKS.entry(workspace.to_path_buf())
+    {
+        let should_remove = {
+            let state = entry.get_mut();
+            if operation.is_exclusive() {
+                state.exclusive = false;
+            } else {
+                state.readers = state.readers.saturating_sub(1);
+            }
+            // Check if we should remove the entry
+            !state.exclusive && state.readers == 0
+        };
+
+        if should_remove {
+            entry.remove();
+        }
+    }
+}
+
+/// Try to acquire a file-based lock for cross-process coordination.
+fn try_acquire_file_lock(
+    lock_path: &Path,
+    operation: IndexOperation,
+) -> Result<(), (String, Option<u32>)> {
+    // Check if existing lock file is valid
+    if let Ok(contents) = std::fs::read_to_string(lock_path)
+        && let Some((op, pid, started)) = parse_lock_file(&contents)
+    {
+        // Check if lock is stale
+        let age = SystemTime::now()
+            .duration_since(started)
+            .unwrap_or(Duration::ZERO);
+
+        if age < operation.stale_timeout() && is_process_alive(pid) {
+            return Err((op, Some(pid)));
+        }
+        // Lock is stale - we can take over
+        tracing::debug!(
+            lock_path = %lock_path.display(),
+            stale_op = %op,
+            stale_pid = pid,
+            age_secs = age.as_secs(),
+            "Taking over stale lock"
+        );
+    }
+
+    // Create parent directory if needed
+    if let Some(parent) = lock_path.parent()
+        && let Err(e) = std::fs::create_dir_all(parent)
+    {
+        tracing::warn!(
+            path = %parent.display(),
+            error = %e,
+            "Failed to create lock directory"
+        );
+    }
+
+    // Write our lock file
+    let contents = format!(
+        "operation={}\npid={}\nstarted={}\nworkspace={}\n",
+        operation.as_str(),
+        std::process::id(),
+        SystemTime::now()
+            .duration_since(UNIX_EPOCH)
+            .unwrap()
+            .as_secs(),
+        lock_path
+            .parent()
+            .and_then(|p| p.file_name())
+            .and_then(|n| n.to_str())
+            .unwrap_or("unknown")
+    );
+
+    std::fs::write(lock_path, contents).map_err(|e| (format!("io_error: {}", e), None))
+}
+
+/// Parse a lock file's contents.
+fn parse_lock_file(contents: &str) -> Option<(String, u32, SystemTime)> {
+    let mut operation = None;
+    let mut pid = None;
+    let mut started = None;
+
+    for line in contents.lines() {
+        if let Some(val) = line.strip_prefix("operation=") {
+            operation = Some(val.to_string());
+        } else if let Some(val) = line.strip_prefix("pid=") {
+            pid = val.parse().ok();
+        } else if let Some(val) = line.strip_prefix("started=")
+            && let Ok(secs) = val.parse::<u64>()
+        {
+            started = Some(UNIX_EPOCH + Duration::from_secs(secs));
+        }
+    }
+
+    match (operation, pid, started) {
+        (Some(op), Some(p), Some(s)) => Some((op, p, s)),
+        _ => None,
+    }
+}
+
+/// Check if a process is still alive.
+#[cfg(unix)]
+fn is_process_alive(pid: u32) -> bool {
+    // kill with signal 0 checks if process exists without sending a signal
+    // Returns 0 if process exists and we have permission to send signals
+    // Returns -1 with ESRCH if process doesn't exist
+    unsafe { libc::kill(pid as libc::pid_t, 0) == 0 }
+}
+
+#[cfg(not(unix))]
+fn is_process_alive(_pid: u32) -> bool {
+    // On non-Unix platforms, rely on timeout-based stale detection
+    true
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use tempfile::tempdir;
+
+    #[test]
+    fn test_exclusive_lock_blocks_exclusive() {
+        let dir = tempdir().unwrap();
+        let workspace = dir.path();
+
+        // Acquire first exclusive lock
+        let guard1 = try_lock(workspace, IndexOperation::Build);
+        assert!(guard1.is_acquired());
+
+        // Second exclusive lock should fail
+        let result2 = try_lock(workspace, IndexOperation::Build);
+        assert!(!result2.is_acquired());
+
+        // Drop first lock
+        drop(guard1);
+
+        // Now second should succeed
+        let guard3 = try_lock(workspace, IndexOperation::Build);
+        assert!(guard3.is_acquired());
+    }
+
+    #[test]
+    fn test_shared_locks_coexist() {
+        let dir = tempdir().unwrap();
+        let workspace = dir.path();
+
+        // Multiple shared locks should work
+        let guard1 = try_lock(workspace, IndexOperation::Load);
+        assert!(guard1.is_acquired());
+
+        let guard2 = try_lock(workspace, IndexOperation::Load);
+        assert!(guard2.is_acquired());
+
+        let guard3 = try_lock(workspace, IndexOperation::Load);
+        assert!(guard3.is_acquired());
+    }
+
+    #[test]
+    fn test_exclusive_blocks_shared() {
+        let dir = tempdir().unwrap();
+        let workspace = dir.path();
+
+        // Acquire exclusive lock
+        let guard1 = try_lock(workspace, IndexOperation::Build);
+        assert!(guard1.is_acquired());
+
+        // Shared lock should fail
+        let result2 = try_lock(workspace, IndexOperation::Load);
+        assert!(!result2.is_acquired());
+    }
+
+    #[test]
+    fn test_shared_blocks_exclusive() {
+        let dir = tempdir().unwrap();
+        let workspace = dir.path();
+
+        // Acquire shared lock
+        let guard1 = try_lock(workspace, IndexOperation::Load);
+        assert!(guard1.is_acquired());
+
+        // Exclusive lock should fail
+        let result2 = try_lock(workspace, IndexOperation::Build);
+        assert!(!result2.is_acquired());
+
+        // Drop shared lock
+        drop(guard1);
+
+        // Now exclusive should succeed
+        let guard3 = try_lock(workspace, IndexOperation::Build);
+        assert!(guard3.is_acquired());
+    }
+
+    #[test]
+    fn test_different_workspaces_independent() {
+        let dir1 = tempdir().unwrap();
+        let dir2 = tempdir().unwrap();
+
+        // Locks on different workspaces should be independent
+        let guard1 = try_lock(dir1.path(), IndexOperation::Build);
+        assert!(guard1.is_acquired());
+
+        let guard2 = try_lock(dir2.path(), IndexOperation::Build);
+        assert!(guard2.is_acquired());
+    }
+
+    #[test]
+    fn test_lock_file_created_for_exclusive() {
+        let dir = tempdir().unwrap();
+        let workspace = dir.path();
+
+        let lock_file = get_lock_file_path(workspace);
+
+        // No lock file initially
+        assert!(!lock_file.exists());
+
+        // Acquire exclusive lock
+        let guard = try_lock(workspace, IndexOperation::Build);
+        assert!(guard.is_acquired());
+
+        // Lock file should exist
+        assert!(lock_file.exists());
+
+        // Check contents
+        let contents = std::fs::read_to_string(&lock_file).unwrap();
+        assert!(contents.contains("operation=build"));
+        assert!(contents.contains(&format!("pid={}", std::process::id())));
+
+        // Drop guard
+        drop(guard);
+
+        // Lock file should be removed
+        assert!(!lock_file.exists());
+    }
+}

crates/codegen/kigi-codebase-graph/src/manager/mod.rs

@@ -0,0 +1,14 @@
+//! Index management: building, caching, locking, and updating.
+
+mod builder;
+pub mod cache;
+pub mod lock;
+
+pub use builder::{IndexBuilder, IndexError, Result};
+pub use cache::{
+    CACHE_FILE_NAME, CacheError, cache_exists, cache_size, get_cache_path, load_index, save_index,
+    save_index_async,
+};
+pub use lock::{
+    IndexOperation, LockResult, WorkspaceLockGuard, is_operation_in_progress, try_lock,
+};

crates/codegen/kigi-codebase-graph/src/navigation.rs

@@ -0,0 +1,844 @@
+//! Location-based navigation APIs for go-to-definition and go-to-references.
+//!
+//! This module provides APIs that take a file path and position (row, column)
+//! and return definition or reference locations.
+
+use std::path::{Path, PathBuf};
+use std::sync::Arc;
+
+use crate::languages::LanguageRegistry;
+use crate::scope_graph::ScopeGraphIndex;
+
+/// Result of a navigation operation (go-to-definition or go-to-references).
+#[derive(Debug, Clone)]
+pub struct NavigationResult {
+    /// The symbol that was found at the query position.
+    pub symbol: String,
+    /// List of locations where the symbol is defined/referenced.
+    pub locations: Vec<Location>,
+}
+
+/// A location in a file.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct Location {
+    /// Path to the file.
+    pub path: String,
+    /// 1-indexed line number.
+    pub line: usize,
+    /// Optional: the matched symbol name (useful for aliases).
+    pub symbol: Option<String>,
+}
+
+impl Location {
+    pub fn new(path: impl Into<String>, line: usize) -> Self {
+        Self {
+            path: path.into(),
+            line,
+            symbol: None,
+        }
+    }
+
+    pub fn with_symbol(path: impl Into<String>, line: usize, symbol: String) -> Self {
+        Self {
+            path: path.into(),
+            line,
+            symbol: Some(symbol),
+        }
+    }
+
+    /// Get the path as a Path reference.
+    pub fn as_path(&self) -> &Path {
+        Path::new(&self.path)
+    }
+}
+
+/// Error type for navigation operations.
+#[derive(Debug)]
+pub enum NavigationError {
+    /// File not found or could not be read.
+    FileNotFound(PathBuf),
+    /// Position is out of bounds for the file.
+    PositionOutOfBounds { row: usize, col: usize },
+    /// No symbol found at the given position.
+    NoSymbolAtPosition { row: usize, col: usize },
+    /// Language not supported for this file type.
+    UnsupportedLanguage(String),
+    /// Parse error.
+    ParseError(String),
+    /// IO error.
+    IoError(std::io::Error),
+}
+
+impl std::fmt::Display for NavigationError {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            NavigationError::FileNotFound(path) => {
+                write!(f, "File not found: {}", path.display())
+            }
+            NavigationError::PositionOutOfBounds { row, col } => {
+                write!(f, "Position out of bounds: {}:{}", row, col)
+            }
+            NavigationError::NoSymbolAtPosition { row, col } => {
+                write!(f, "No symbol found at position {}:{}", row, col)
+            }
+            NavigationError::UnsupportedLanguage(ext) => {
+                write!(f, "Unsupported language: {}", ext)
+            }
+            NavigationError::ParseError(msg) => write!(f, "Parse error: {}", msg),
+            NavigationError::IoError(e) => write!(f, "IO error: {}", e),
+        }
+    }
+}
+
+impl std::error::Error for NavigationError {}
+
+impl From<std::io::Error> for NavigationError {
+    fn from(e: std::io::Error) -> Self {
+        NavigationError::IoError(e)
+    }
+}
+
+/// Navigator provides location-based code navigation.
+///
+/// It wraps a ScopeGraphIndex and provides methods to navigate code
+/// based on file path and position (row, column).
+pub struct Navigator {
+    index: Arc<ScopeGraphIndex>,
+    registry: LanguageRegistry,
+}
+
+impl Navigator {
+    /// Create a new Navigator backed by a shared index.
+    ///
+    /// Accepts anything that converts into `Arc<ScopeGraphIndex>`, so both
+    /// owned and already-shared indexes work without extra wrapping:
+    ///
+    /// ```rust,ignore
+    /// // From an owned index (e.g. IndexBuilder)
+    /// let navigator = Navigator::new(index);
+    ///
+    /// // From a shared snapshot (zero-cost)
+    /// let snapshot = handle.get_snapshot()?;
+    /// let navigator = Navigator::new(snapshot);
+    /// ```
+    pub fn new(index: impl Into<Arc<ScopeGraphIndex>>) -> Self {
+        Self {
+            index: index.into(),
+            registry: LanguageRegistry::new(),
+        }
+    }
+
+    /// Get a reference to the underlying index.
+    pub fn index(&self) -> &ScopeGraphIndex {
+        &self.index
+    }
+
+    /// Get a mutable reference to the underlying index.
+    ///
+    /// Uses copy-on-write: if other `Arc` clones of the index exist, the
+    /// index is cloned before returning the mutable reference.
+    pub fn index_mut(&mut self) -> &mut ScopeGraphIndex {
+        Arc::make_mut(&mut self.index)
+    }
+
+    /// Get the symbol at the given file path and position.
+    ///
+    /// # Arguments
+    /// * `file_path` - Path to the file
+    /// * `row` - 1-indexed line number
+    /// * `col` - 1-indexed column number
+    ///
+    /// # Returns
+    /// The symbol name at the given position.
+    pub fn get_symbol_at_position(
+        &self,
+        file_path: &Path,
+        row: usize,
+        col: usize,
+    ) -> Result<String, NavigationError> {
+        // Validate position (1-indexed)
+        if row == 0 || col == 0 {
+            return Err(NavigationError::PositionOutOfBounds { row, col });
+        }
+
+        let content = std::fs::read(file_path)
+            .map_err(|_| NavigationError::FileNotFound(file_path.to_path_buf()))?;
+
+        // Get the language config
+        let lang_config = self.registry.for_file_path(file_path).ok_or_else(|| {
+            NavigationError::UnsupportedLanguage(
+                file_path
+                    .extension()
+                    .and_then(|e| e.to_str())
+                    .unwrap_or("unknown")
+                    .to_string(),
+            )
+        })?;
+
+        // Parse the file
+        let mut parser = tree_sitter::Parser::new();
+        parser
+            .set_language(&lang_config.language())
+            .map_err(|e| NavigationError::ParseError(format!("Failed to set language: {}", e)))?;
+
+        let tree = parser
+            .parse(&content, None)
+            .ok_or_else(|| NavigationError::ParseError("Failed to parse file".to_string()))?;
+
+        // Convert 1-indexed row/col to 0-indexed for tree-sitter
+        let point = tree_sitter::Point::new(row - 1, col - 1);
+
+        // Find the node at the position
+        let root = tree.root_node();
+        let node = find_smallest_named_node_at_point(root, point);
+
+        match node {
+            Some(n) => {
+                let text = std::str::from_utf8(&content[n.byte_range()])
+                    .map_err(|_| NavigationError::ParseError("Invalid UTF-8".to_string()))?;
+                Ok(text.to_string())
+            }
+            None => Err(NavigationError::NoSymbolAtPosition { row, col }),
+        }
+    }
+
+    /// Go to definition for the symbol at the given position.
+    ///
+    /// # Arguments
+    /// * `file_path` - Path to the file
+    /// * `row` - 1-indexed line number
+    /// * `col` - 1-indexed column number
+    ///
+    /// # Returns
+    /// NavigationResult containing the symbol and its definition locations.
+    pub fn goto_definition(
+        &self,
+        file_path: &Path,
+        row: usize,
+        col: usize,
+    ) -> Result<NavigationResult, NavigationError> {
+        let symbol = self.get_symbol_at_position(file_path, row, col)?;
+
+        // Look up definitions in the index
+        let defs =
+            self.index
+                .find_definitions_smart(&symbol, Some(file_path), Some(&self.registry));
+
+        let locations: Vec<Location> = defs
+            .into_iter()
+            .map(|(path, line)| Location::new(path, line))
+            .collect();
+
+        Ok(NavigationResult { symbol, locations })
+    }
+
+    /// Go to references for the symbol at the given position.
+    ///
+    /// This first resolves the symbol to its definition, then finds all references.
+    ///
+    /// # Arguments
+    /// * `file_path` - Path to the file
+    /// * `row` - 1-indexed line number  
+    /// * `col` - 1-indexed column number
+    /// * `include_definition` - Whether to include the definition location in results
+    ///
+    /// # Returns
+    /// NavigationResult containing the symbol and its reference locations.
+    pub fn goto_references(
+        &self,
+        file_path: &Path,
+        row: usize,
+        col: usize,
+        include_definition: bool,
+    ) -> Result<NavigationResult, NavigationError> {
+        let symbol = self.get_symbol_at_position(file_path, row, col)?;
+
+        // Get references (includes alias resolution)
+        let refs = self
+            .index
+            .find_references_smart(&symbol, Some(file_path), Some(&self.registry));
+
+        let mut locations: Vec<Location> = refs
+            .into_iter()
+            .map(|(sym, path, line)| Location::with_symbol(path, line, sym))
+            .collect();
+
+        // Optionally include definition locations
+        if include_definition {
+            let defs =
+                self.index
+                    .find_definitions_smart(&symbol, Some(file_path), Some(&self.registry));
+
+            for (path, line) in defs {
+                let loc = Location::new(path, line);
+                if !locations
+                    .iter()
+                    .any(|l| l.path == loc.path && l.line == loc.line)
+                {
+                    locations.insert(0, loc);
+                }
+            }
+        }
+
+        Ok(NavigationResult { symbol, locations })
+    }
+
+    /// Go to definition by symbol name directly (without position lookup).
+    pub fn goto_definition_by_name(
+        &self,
+        symbol: &str,
+        context_file: Option<&Path>,
+    ) -> NavigationResult {
+        let defs = self
+            .index
+            .find_definitions_smart(symbol, context_file, Some(&self.registry));
+
+        let locations: Vec<Location> = defs
+            .into_iter()
+            .map(|(path, line)| Location::new(path, line))
+            .collect();
+
+        NavigationResult {
+            symbol: symbol.to_string(),
+            locations,
+        }
+    }
+
+    /// Go to references by symbol name directly (without position lookup).
+    pub fn goto_references_by_name(
+        &self,
+        symbol: &str,
+        context_file: Option<&Path>,
+        include_definition: bool,
+    ) -> NavigationResult {
+        let refs = self
+            .index
+            .find_references_smart(symbol, context_file, Some(&self.registry));
+
+        let mut locations: Vec<Location> = refs
+            .into_iter()
+            .map(|(sym, path, line)| Location::with_symbol(path, line, sym))
+            .collect();
+
+        if include_definition {
+            let defs =
+                self.index
+                    .find_definitions_smart(symbol, context_file, Some(&self.registry));
+
+            for (path, line) in defs {
+                let loc = Location::new(path, line);
+                if !locations
+                    .iter()
+                    .any(|l| l.path == loc.path && l.line == loc.line)
+                {
+                    locations.insert(0, loc);
+                }
+            }
+        }
+
+        NavigationResult {
+            symbol: symbol.to_string(),
+            locations,
+        }
+    }
+}
+
+/// Find the smallest named node that contains the given point.
+fn find_smallest_named_node_at_point(
+    node: tree_sitter::Node<'_>,
+    point: tree_sitter::Point,
+) -> Option<tree_sitter::Node<'_>> {
+    // Check if point is within this node
+    if point < node.start_position() || point > node.end_position() {
+        return None;
+    }
+
+    // Try to find a smaller child node that contains the point
+    let mut cursor = node.walk();
+    for child in node.children(&mut cursor) {
+        if let Some(found) = find_smallest_named_node_at_point(child, point) {
+            // Prefer named nodes that look like identifiers
+            if is_identifier_like(&found) {
+                return Some(found);
+            }
+            // Keep searching for a better match
+            if is_identifier_like(&child) {
+                return Some(child);
+            }
+            return Some(found);
+        }
+    }
+
+    // No smaller child contains the point, return this node if it's identifier-like
+    if is_identifier_like(&node) {
+        Some(node)
+    } else {
+        None
+    }
+}
+
+/// Check if a node looks like an identifier.
+fn is_identifier_like(node: &tree_sitter::Node<'_>) -> bool {
+    let kind = node.kind();
+    matches!(
+        kind,
+        "identifier"
+            | "type_identifier"
+            | "property_identifier"
+            | "field_identifier"
+            | "shorthand_property_identifier"
+            | "shorthand_property_identifier_pattern"
+            | "attribute"           // Python
+            | "package_identifier" // Go
+    )
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::IndexBuilder;
+    use std::fs::File;
+    use std::io::Write;
+    use tempfile::tempdir;
+
+    #[test]
+    fn test_get_symbol_at_position() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.rs");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(file, "fn hello_world() {{").unwrap();
+        writeln!(file, "    println!(\"Hello\");").unwrap();
+        writeln!(file, "}}").unwrap();
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Get symbol at "hello_world" (row 1, col 4)
+        let symbol = navigator.get_symbol_at_position(&file_path, 1, 4).unwrap();
+        assert_eq!(symbol, "hello_world");
+    }
+
+    #[test]
+    fn test_typescript_for_of_array_destructuring() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"const myMap = new Map<string, {{ sessionId: string }}>();
+
+function test(sessionId: string) {{
+  for (const [toolCallId, request] of myMap) {{
+    if (request.sessionId !== sessionId) continue;
+    console.log(toolCallId);
+  }}
+}}"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'request' is found as a definition at line 4
+        let def_result = navigator.goto_definition_by_name("request", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "request should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 4,
+            "request should be defined on line 4"
+        );
+
+        // Test that 'toolCallId' is also found as a definition at line 4
+        let def_result2 = navigator.goto_definition_by_name("toolCallId", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "toolCallId should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 4,
+            "toolCallId should be defined on line 4"
+        );
+
+        // Test that references to 'request' are found
+        let ref_result = navigator.goto_references_by_name("request", Some(&file_path), false);
+        assert!(
+            !ref_result.locations.is_empty(),
+            "request should have references"
+        );
+        // Check that line 5 reference is found (request.sessionId)
+        let line5_ref = ref_result.locations.iter().any(|loc| loc.line == 5);
+        assert!(line5_ref, "request should be referenced on line 5");
+    }
+
+    #[test]
+    fn test_typescript_for_of_object_destructuring() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"const items = [{{ name: "a", value: 1 }}];
+
+for (const {{ name, value }} of items) {{
+  console.log(name, value);
+}}"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'name' is found as a definition at line 3
+        let def_result = navigator.goto_definition_by_name("name", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "name should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 3,
+            "name should be defined on line 3"
+        );
+
+        // Test that 'value' is found as a definition at line 3
+        let def_result2 = navigator.goto_definition_by_name("value", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "value should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 3,
+            "value should be defined on line 3"
+        );
+    }
+
+    #[test]
+    fn test_typescript_regular_array_destructuring() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"const arr = [1, 2, 3];
+const [first, second, third] = arr;
+console.log(first, second);"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'first' is found as a definition at line 2
+        let def_result = navigator.goto_definition_by_name("first", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "first should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 2,
+            "first should be defined on line 2"
+        );
+
+        // Test that 'second' is found as a definition at line 2
+        let def_result2 = navigator.goto_definition_by_name("second", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "second should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 2,
+            "second should be defined on line 2"
+        );
+    }
+
+    #[test]
+    fn test_typescript_regular_object_destructuring() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"const obj = {{ foo: 1, bar: 2 }};
+const {{ foo, bar }} = obj;
+console.log(foo, bar);"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'foo' is found as a definition at line 2
+        let def_result = navigator.goto_definition_by_name("foo", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "foo should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 2,
+            "foo should be defined on line 2"
+        );
+
+        // Test that 'bar' is found as a definition at line 2
+        let def_result2 = navigator.goto_definition_by_name("bar", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "bar should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 2,
+            "bar should be defined on line 2"
+        );
+    }
+
+    #[test]
+    fn test_typescript_member_expression_reference() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"const myObject = {{ value: 42 }};
+const result = myObject.value;"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'myObject' is referenced on line 2 (myObject.value)
+        let ref_result = navigator.goto_references_by_name("myObject", Some(&file_path), false);
+        assert!(
+            !ref_result.locations.is_empty(),
+            "myObject should have references"
+        );
+        let line2_ref = ref_result.locations.iter().any(|loc| loc.line == 2);
+        assert!(line2_ref, "myObject should be referenced on line 2");
+    }
+
+    #[test]
+    fn test_typescript_function_parameter_destructuring() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"function process([first, second]: number[], {{ name }}: {{ name: string }}) {{
+  console.log(first, second, name);
+}}"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'first' is found as a definition at line 1
+        let def_result = navigator.goto_definition_by_name("first", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "first should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 1,
+            "first should be defined on line 1"
+        );
+
+        // Test that 'name' is found as a definition at line 1
+        let def_result2 = navigator.goto_definition_by_name("name", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "name should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 1,
+            "name should be defined on line 1"
+        );
+    }
+
+    #[test]
+    fn test_typescript_simple_function_parameters() {
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("test.ts");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"function greet(name: string, age: number) {{
+  console.log(name, age);
+}}"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'name' is found as a definition at line 1
+        let def_result = navigator.goto_definition_by_name("name", Some(&file_path));
+        assert!(
+            !def_result.locations.is_empty(),
+            "name should be found as a definition"
+        );
+        assert_eq!(
+            def_result.locations[0].line, 1,
+            "name should be defined on line 1"
+        );
+
+        // Test that 'age' is found as a definition at line 1
+        let def_result2 = navigator.goto_definition_by_name("age", Some(&file_path));
+        assert!(
+            !def_result2.locations.is_empty(),
+            "age should be found as a definition"
+        );
+        assert_eq!(
+            def_result2.locations[0].line, 1,
+            "age should be defined on line 1"
+        );
+    }
+
+    #[test]
+    fn test_typescript_react_dependency_array_references() {
+        // Test that identifiers in React hook dependency arrays are captured as references
+        // Regression test for dependency-array reference capture.
+        let dir = tempdir().unwrap();
+        let file_path = dir.path().join("component.tsx");
+
+        let mut file = File::create(&file_path).unwrap();
+        writeln!(
+            file,
+            r#"import {{ useCallback, useEffect }} from 'react';
+
+function FileTreeTab({{ basePath, onFileSelect }}) {{
+  const fileTree = useFileTree();
+  
+  const loadDirectory = useCallback(
+    (path: string) => {{
+      return fileTree.listFiles(path);
+    }},
+    [fileTree],
+  );
+
+  const handleOpenPath = useCallback(
+    (path: string) => {{
+      loadDirectory(path);
+    }},
+    [loadDirectory, fileTree],
+  );
+
+  const handleFileSelect = useCallback(
+    (file) => {{
+      fileTree.setSelectedPath(file.absPath);
+      onFileSelect(file.absPath);
+    }},
+    [fileTree, onFileSelect],
+  );
+
+  useEffect(() => {{
+    loadDirectory(basePath);
+  }}, [basePath, loadDirectory]);
+
+  return null;
+}}"#
+        )
+        .unwrap();
+        drop(file);
+
+        let index = IndexBuilder::new().build(dir.path()).unwrap();
+        let navigator = Navigator::new(index);
+
+        // Test that 'fileTree' is found in dependency arrays
+        let ref_result = navigator.goto_references_by_name("fileTree", Some(&file_path), false);
+        assert!(
+            !ref_result.locations.is_empty(),
+            "fileTree should have references"
+        );
+
+        // Check references in dependency arrays:
+        // Line 10: [fileTree]
+        // Line 17: [loadDirectory, fileTree]
+        // Line 25: [fileTree, onFileSelect]
+        let dep_array_lines: Vec<usize> = ref_result
+            .locations
+            .iter()
+            .filter(|loc| loc.line == 10 || loc.line == 17 || loc.line == 25)
+            .map(|loc| loc.line)
+            .collect();
+
+        assert!(
+            dep_array_lines.contains(&10),
+            "fileTree should be referenced on line 10 (first dependency array)"
+        );
+        assert!(
+            dep_array_lines.contains(&17),
+            "fileTree should be referenced on line 17 (second dependency array)"
+        );
+        assert!(
+            dep_array_lines.contains(&25),
+            "fileTree should be referenced on line 25 (third dependency array)"
+        );
+
+        // Test that 'loadDirectory' is found in dependency arrays
+        let ref_result2 =
+            navigator.goto_references_by_name("loadDirectory", Some(&file_path), false);
+        assert!(
+            !ref_result2.locations.is_empty(),
+            "loadDirectory should have references"
+        );
+
+        // Line 17: [loadDirectory, fileTree]
+        // Line 30: [basePath, loadDirectory]
+        let load_dir_refs: Vec<usize> = ref_result2
+            .locations
+            .iter()
+            .filter(|loc| loc.line == 17 || loc.line == 30)
+            .map(|loc| loc.line)
+            .collect();
+
+        assert!(
+            load_dir_refs.contains(&17),
+            "loadDirectory should be referenced on line 17"
+        );
+        assert!(
+            load_dir_refs.contains(&30),
+            "loadDirectory should be referenced on line 30"
+        );
+
+        // Test that 'onFileSelect' is found in dependency array
+        let ref_result3 =
+            navigator.goto_references_by_name("onFileSelect", Some(&file_path), false);
+        assert!(
+            !ref_result3.locations.is_empty(),
+            "onFileSelect should have references"
+        );
+        let on_file_select_line25 = ref_result3.locations.iter().any(|loc| loc.line == 25);
+        assert!(
+            on_file_select_line25,
+            "onFileSelect should be referenced on line 25 (dependency array)"
+        );
+
+        // Test that 'basePath' is found in dependency array
+        let ref_result4 = navigator.goto_references_by_name("basePath", Some(&file_path), false);
+        assert!(
+            !ref_result4.locations.is_empty(),
+            "basePath should have references"
+        );
+        let base_path_line30 = ref_result4.locations.iter().any(|loc| loc.line == 30);
+        assert!(
+            base_path_line30,
+            "basePath should be referenced on line 30 (useEffect dependency array)"
+        );
+    }
+}

crates/codegen/kigi-codebase-graph/src/scope_graph/edges.rs

@@ -0,0 +1,22 @@
+//! Edge types for the ScopeGraph.
+
+use serde::{Deserialize, Serialize};
+
+/// Describes the relation between two nodes in the ScopeGraph.
+#[derive(Serialize, Deserialize, PartialEq, Eq, Copy, Clone, Debug)]
+pub enum EdgeKind {
+    /// The edge weight from a nested scope to its parent scope.
+    ScopeToScope,
+
+    /// The edge weight from a definition to its definition scope.
+    DefToScope,
+
+    /// The edge weight from an import to its definition scope.
+    ImportToScope,
+
+    /// The edge weight from a reference to its definition.
+    RefToDef,
+
+    /// The edge weight from a reference to its import.
+    RefToImport,
+}

crates/codegen/kigi-codebase-graph/src/scope_graph/mod.rs

@@ -0,0 +1,38 @@
+//! ScopeGraph module for per-file symbol tracking.
+//!
+//! A ScopeGraph represents the symbols (definitions, references, imports) and their
+//! relationships within a single source file.
+
+pub mod edges;
+pub mod graph;
+pub mod nodes;
+
+pub use edges::EdgeKind;
+pub use graph::{
+    NodeIndex, QueryVersion, ScopeGraph, ScopeGraphIndex, ScopeStack, Snippet,
+    extract_symbols_fast, scope_graph_from_definitions_query,
+};
+pub use nodes::{LocalDef, LocalImport, LocalScope, NodeKind, Reference, Symbol, SymbolId};
+
+use crate::languages::TSLanguageConfig;
+
+/// Result of building a scope graph, including alias pairs.
+pub struct ScopeGraphResult {
+    /// The scope graph for the file.
+    pub graph: ScopeGraph,
+    /// Alias pairs: (alias_name, original_name).
+    pub aliases: Vec<(String, String)>,
+}
+
+/// Build a ScopeGraph from tree-sitter query and source.
+///
+/// This is a convenience wrapper around `scope_graph_from_definitions_query`.
+pub fn build_scope_graph(
+    query: &tree_sitter::Query,
+    root_node: tree_sitter::Node<'_>,
+    src: &[u8],
+    language: &TSLanguageConfig,
+) -> ScopeGraphResult {
+    let (graph, aliases) = scope_graph_from_definitions_query(query, root_node, src, language);
+    ScopeGraphResult { graph, aliases }
+}

crates/codegen/kigi-codebase-graph/src/scope_graph/nodes.rs

@@ -0,0 +1,180 @@
+//! Node types for the ScopeGraph.
+
+use std::sync::Arc;
+
+use serde::{Deserialize, Serialize};
+
+use crate::types::Range;
+
+/// A symbol extracted from the code.
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Symbol {
+    /// The kind/type of symbol (e.g., "function", "class", etc.)
+    pub kind: Arc<str>,
+    /// The range where the symbol appears.
+    pub range: Range,
+}
+
+impl Symbol {
+    /// Create a new symbol.
+    pub fn new(kind: Arc<str>, range: Range) -> Self {
+        Self { kind, range }
+    }
+}
+
+/// An opaque identifier for every symbol in a language.
+#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Eq, Hash)]
+pub struct SymbolId {
+    /// Index into the namespace array.
+    pub namespace_idx: usize,
+    /// Index within the namespace.
+    pub symbol_idx: usize,
+}
+
+impl SymbolId {
+    /// Create a new symbol ID.
+    pub fn new(namespace_idx: usize, symbol_idx: usize) -> Self {
+        Self {
+            namespace_idx,
+            symbol_idx,
+        }
+    }
+
+    /// Get the symbol name from the namespaces.
+    pub fn name<'a>(&self, namespaces: &'a [Vec<String>]) -> Option<&'a str> {
+        namespaces
+            .get(self.namespace_idx)
+            .and_then(|ns| ns.get(self.symbol_idx))
+            .map(|s| s.as_str())
+    }
+}
+
+/// A local scope in the source code.
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Hash, Eq)]
+pub struct LocalScope {
+    /// The range of this scope.
+    pub range: Range,
+}
+
+impl LocalScope {
+    /// Create a new local scope.
+    pub fn new(range: Range) -> Self {
+        Self { range }
+    }
+}
+
+/// A local definition in the source code.
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq, Hash)]
+pub struct LocalDef {
+    /// The range of the identifier being defined.
+    pub range: Range,
+    /// Optional symbol ID for type-aware resolution.
+    pub symbol_id: Option<SymbolId>,
+    /// The scope where this definition is visible.
+    pub scope: LocalScope,
+}
+
+impl LocalDef {
+    /// Create a new local definition.
+    pub fn new(range: Range, symbol_id: Option<SymbolId>, scope: LocalScope) -> Self {
+        Self {
+            range,
+            symbol_id,
+            scope,
+        }
+    }
+
+    /// Get the name of this definition from source bytes.
+    pub fn name<'a>(&self, src: &'a [u8]) -> &'a [u8] {
+        &src[self.range.start_byte()..self.range.end_byte()]
+    }
+
+    /// Get the scope range.
+    pub fn scope_range(&self) -> &Range {
+        &self.scope.range
+    }
+}
+
+/// A local import in the source code.
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq, Hash)]
+pub struct LocalImport {
+    /// The range of the import identifier.
+    pub range: Range,
+}
+
+impl LocalImport {
+    /// Create a new local import.
+    pub fn new(range: Range) -> Self {
+        Self { range }
+    }
+
+    /// Get the name of this import from source bytes.
+    pub fn name<'a>(&self, src: &'a [u8]) -> &'a [u8] {
+        &src[self.range.start_byte()..self.range.end_byte()]
+    }
+}
+
+/// A reference to a symbol in the source code.
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct Reference {
+    /// The range of the reference.
+    pub range: Range,
+    /// Optional symbol ID for type-aware resolution.
+    pub symbol_id: Option<SymbolId>,
+}
+
+impl Reference {
+    /// Create a new reference.
+    pub fn new(range: Range, symbol_id: Option<SymbolId>) -> Self {
+        Self { range, symbol_id }
+    }
+
+    /// Get the name of this reference from source bytes.
+    pub fn name<'a>(&self, src: &'a [u8]) -> &'a [u8] {
+        &src[self.range.start_byte()..self.range.end_byte()]
+    }
+}
+
+/// The type of a node in the ScopeGraph.
+#[derive(Serialize, Deserialize, Debug, Clone)]
+pub enum NodeKind {
+    /// A scope node.
+    Scope(LocalScope),
+
+    /// A definition node.
+    Def(LocalDef),
+
+    /// An import node.
+    Import(LocalImport),
+
+    /// A reference node.
+    Ref(Reference),
+}
+
+impl NodeKind {
+    /// Construct a scope node from a range.
+    pub fn scope(range: Range) -> Self {
+        Self::Scope(LocalScope::new(range))
+    }
+
+    /// Produce the range spanned by this node.
+    pub fn range(&self) -> Range {
+        match self {
+            Self::Scope(l) => l.range,
+            // For definitions, return the scope range to capture the full context
+            Self::Def(d) => d.scope.range,
+            Self::Ref(r) => r.range,
+            Self::Import(i) => i.range,
+        }
+    }
+
+    /// Get the identifier range (the actual symbol location).
+    pub fn identifier_range(&self) -> Range {
+        match self {
+            Self::Scope(l) => l.range,
+            Self::Def(d) => d.range,
+            Self::Ref(r) => r.range,
+            Self::Import(i) => i.range,
+        }
+    }
+}

crates/codegen/kigi-codebase-graph/src/types/file_event.rs

@@ -0,0 +1,78 @@
+//! File change events for live index updates.
+
+use std::path::PathBuf;
+
+/// Events that can trigger index updates.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum FileEvent {
+    /// A new file was created.
+    Created {
+        /// Path to the new file.
+        path: PathBuf,
+    },
+
+    /// An existing file was modified.
+    Modified {
+        /// Path to the modified file.
+        path: PathBuf,
+    },
+
+    /// A file was deleted.
+    Deleted {
+        /// Path to the deleted file.
+        path: PathBuf,
+    },
+
+    /// A file was renamed/moved.
+    Renamed {
+        /// Original path.
+        from: PathBuf,
+        /// New path.
+        to: PathBuf,
+    },
+}
+
+impl FileEvent {
+    /// Get the primary path associated with this event.
+    pub fn path(&self) -> &PathBuf {
+        match self {
+            FileEvent::Created { path } => path,
+            FileEvent::Modified { path } => path,
+            FileEvent::Deleted { path } => path,
+            FileEvent::Renamed { to, .. } => to,
+        }
+    }
+
+    /// Check if this event requires reparsing the file content.
+    pub fn requires_reparse(&self) -> bool {
+        match self {
+            FileEvent::Created { .. } => true,
+            FileEvent::Modified { .. } => true,
+            FileEvent::Deleted { .. } => false,
+            FileEvent::Renamed { .. } => false, // Only path update needed
+        }
+    }
+
+    /// Check if this event affects an existing indexed file.
+    pub fn affects_existing(&self) -> bool {
+        match self {
+            FileEvent::Created { .. } => false,
+            FileEvent::Modified { .. } => true,
+            FileEvent::Deleted { .. } => true,
+            FileEvent::Renamed { .. } => true,
+        }
+    }
+}
+
+impl std::fmt::Display for FileEvent {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            FileEvent::Created { path } => write!(f, "Created: {}", path.display()),
+            FileEvent::Modified { path } => write!(f, "Modified: {}", path.display()),
+            FileEvent::Deleted { path } => write!(f, "Deleted: {}", path.display()),
+            FileEvent::Renamed { from, to } => {
+                write!(f, "Renamed: {} -> {}", from.display(), to.display())
+            }
+        }
+    }
+}

crates/codegen/kigi-codebase-graph/src/types/location.rs

@@ -0,0 +1,108 @@
+//! Location type for query results.
+//!
+//! Location uses 1-indexed line and column numbers for LSP compatibility.
+
+use std::path::PathBuf;
+
+use serde::{Deserialize, Serialize};
+
+use super::Range;
+
+/// A location in the codebase, used for query results.
+///
+/// Following LSP protocol conventions:
+/// - `file_path`: Absolute path to the file
+/// - `line`: 1-indexed line number
+/// - `column`: 1-indexed column number
+/// - `range`: Full range information (0-indexed internally)
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq, Hash)]
+#[serde(rename_all = "camelCase")]
+pub struct Location {
+    /// File path (absolute)
+    pub file_path: PathBuf,
+    /// Line number (1-indexed)
+    pub line: usize,
+    /// Column number (1-indexed)
+    pub column: usize,
+    /// Full range information (0-indexed internally)
+    pub range: Range,
+}
+
+impl Location {
+    /// Create a new location with 1-indexed line and column.
+    pub fn new(file_path: PathBuf, line: usize, column: usize, range: Range) -> Self {
+        Self {
+            file_path,
+            line,
+            column,
+            range,
+        }
+    }
+
+    /// Create a location from a range (automatically converts to 1-indexed).
+    pub fn from_range(file_path: PathBuf, range: Range) -> Self {
+        Self {
+            file_path,
+            line: range.start_line_1indexed(),
+            column: range.start_column_1indexed(),
+            range,
+        }
+    }
+
+    /// Get the file path.
+    pub fn file_path(&self) -> &PathBuf {
+        &self.file_path
+    }
+
+    /// Alias for file_path() - for compatibility.
+    pub fn path(&self) -> &PathBuf {
+        &self.file_path
+    }
+
+    /// Get the 1-indexed line number.
+    pub fn line(&self) -> usize {
+        self.line
+    }
+
+    /// Get the 1-indexed column number.
+    pub fn column(&self) -> usize {
+        self.column
+    }
+
+    /// Get the range (0-indexed internally).
+    pub fn range(&self) -> &Range {
+        &self.range
+    }
+
+    /// Get the file extension, if any.
+    pub fn extension(&self) -> Option<&str> {
+        self.file_path.extension().and_then(|e| e.to_str())
+    }
+
+    /// Get the parent directory of the file.
+    pub fn parent_dir(&self) -> Option<&std::path::Path> {
+        self.file_path.parent()
+    }
+
+    /// Get the 0-indexed line number (for internal use).
+    pub fn line_0indexed(&self) -> usize {
+        self.line.saturating_sub(1)
+    }
+
+    /// Get the 0-indexed column number (for internal use).
+    pub fn column_0indexed(&self) -> usize {
+        self.column.saturating_sub(1)
+    }
+}
+
+impl std::fmt::Display for Location {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(
+            f,
+            "{}:{}:{}",
+            self.file_path.display(),
+            self.line,
+            self.column
+        )
+    }
+}

crates/codegen/kigi-codebase-graph/src/types/mod.rs

@@ -0,0 +1,119 @@
+//! Core types for the goto_index crate.
+
+use std::sync::Arc;
+
+mod file_event;
+mod location;
+mod range;
+
+pub use file_event::FileEvent;
+pub use location::Location;
+pub use range::{Position, Range};
+
+/// Statistics about an index.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct IndexStats {
+    /// Number of files indexed
+    pub files: usize,
+    /// Number of symbol definitions
+    pub definitions: usize,
+    /// Number of symbol references
+    pub references: usize,
+}
+
+impl IndexStats {
+    /// Create new index stats.
+    pub fn new(files: usize, definitions: usize, references: usize) -> Self {
+        Self {
+            files,
+            definitions,
+            references,
+        }
+    }
+}
+
+/// A symbol with its line number (1-indexed).
+/// Uses Arc<str> to avoid extra allocation when merging into index.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct SymbolOccurrence {
+    /// The symbol name
+    pub name: Arc<str>,
+    /// Line number (1-indexed)
+    pub line: usize,
+}
+
+impl SymbolOccurrence {
+    /// Create a new symbol occurrence.
+    pub fn new(name: Arc<str>, line: usize) -> Self {
+        Self { name, line }
+    }
+}
+
+/// An alias mapping (alias_name -> original_name).
+/// Uses Arc<str> to avoid extra allocation when merging into index.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct SymbolAlias {
+    /// The alias name (e.g., imported as)
+    pub alias: Arc<str>,
+    /// The original symbol name
+    pub original: Arc<str>,
+}
+
+impl SymbolAlias {
+    /// Create a new symbol alias.
+    pub fn new(alias: Arc<str>, original: Arc<str>) -> Self {
+        Self { alias, original }
+    }
+}
+
+/// File metadata for staleness detection.
+///
+/// Stores size and modification time to quickly detect if a file has changed
+/// without reading its contents.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+pub struct FileMeta {
+    /// File size in bytes
+    pub size: u64,
+    /// Modification time as seconds since UNIX epoch
+    pub mtime_secs: i64,
+    /// Modification time nanoseconds component
+    pub mtime_nanos: u32,
+}
+
+impl FileMeta {
+    /// Create new file metadata.
+    pub fn new(size: u64, mtime_secs: i64, mtime_nanos: u32) -> Self {
+        Self {
+            size,
+            mtime_secs,
+            mtime_nanos,
+        }
+    }
+
+    /// Create file metadata from std::fs::Metadata.
+    pub fn from_metadata(meta: &std::fs::Metadata) -> Self {
+        let size = meta.len();
+        let (mtime_secs, mtime_nanos) = meta
+            .modified()
+            .ok()
+            .and_then(|t| t.duration_since(std::time::UNIX_EPOCH).ok())
+            .map(|d| (d.as_secs() as i64, d.subsec_nanos()))
+            .unwrap_or((0, 0));
+        Self {
+            size,
+            mtime_secs,
+            mtime_nanos,
+        }
+    }
+
+    /// Check if the file has changed compared to current filesystem state.
+    pub fn is_stale(&self, path: &std::path::Path) -> bool {
+        match std::fs::metadata(path) {
+            Ok(meta) => {
+                let current = Self::from_metadata(&meta);
+                *self != current
+            }
+            Err(_) => true, // File deleted or inaccessible
+        }
+    }
+}

crates/codegen/kigi-codebase-graph/src/types/range.rs

@@ -0,0 +1,365 @@
+//! Position and Range types for representing source code locations.
+//!
+//! These types follow LSP conventions:
+//! - Internally stored as 0-indexed (tree-sitter compatible)
+//! - Public API provides both 0-indexed and 1-indexed accessors
+
+use serde::{Deserialize, Serialize};
+
+/// A position in a source file.
+///
+/// Positions are stored as 0-indexed internally (tree-sitter compatible).
+/// Use `line_1indexed()` and `column_1indexed()` for display/LSP output.
+#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Eq, Hash, Default)]
+#[serde(rename_all = "camelCase")]
+pub struct Position {
+    /// Line number (0-indexed)
+    line: usize,
+    /// Character/column number (0-indexed)
+    character: usize,
+    /// Byte offset from start of file
+    byte_offset: usize,
+}
+
+impl Position {
+    /// Create a new position (0-indexed).
+    pub fn new(line: usize, character: usize, byte_offset: usize) -> Self {
+        Self {
+            line,
+            character,
+            byte_offset,
+        }
+    }
+
+    /// Create a position from a tree-sitter point.
+    pub fn from_tree_sitter_point(point: &tree_sitter::Point, byte_offset: usize) -> Self {
+        Self {
+            line: point.row,
+            character: point.column,
+            byte_offset,
+        }
+    }
+
+    /// Convert to tree-sitter Point.
+    pub fn to_tree_sitter(&self) -> tree_sitter::Point {
+        tree_sitter::Point::new(self.line, self.character)
+    }
+
+    /// Get the line number (0-indexed).
+    pub fn line(&self) -> usize {
+        self.line
+    }
+
+    /// Get the line number (1-indexed, for LSP/display).
+    pub fn line_1indexed(&self) -> usize {
+        self.line + 1
+    }
+
+    /// Get the column/character number (0-indexed).
+    pub fn column(&self) -> usize {
+        self.character
+    }
+
+    /// Get the column/character number (1-indexed, for LSP/display).
+    pub fn column_1indexed(&self) -> usize {
+        self.character + 1
+    }
+
+    /// Alias for column() - matches LSP terminology.
+    pub fn character(&self) -> usize {
+        self.character
+    }
+
+    /// Get the byte offset.
+    pub fn byte_offset(&self) -> usize {
+        self.byte_offset
+    }
+
+    /// Get the byte offset (alias).
+    pub fn to_byte_offset(&self) -> usize {
+        self.byte_offset
+    }
+
+    /// Set the byte offset.
+    pub fn set_byte_offset(&mut self, byte_offset: usize) {
+        self.byte_offset = byte_offset;
+    }
+
+    /// Check if this position is before or at another position.
+    pub fn before_other(&self, other: &Position) -> bool {
+        self.line < other.line || (self.line == other.line && self.character <= other.character)
+    }
+
+    /// Check if this position is after or at another position.
+    pub fn after_other(&self, other: &Position) -> bool {
+        self.line > other.line || (self.line == other.line && self.character >= other.character)
+    }
+
+    /// Create a position from a byte offset and line end indices.
+    pub fn from_byte(byte: usize, line_end_indices: &[u32]) -> Self {
+        let line = line_end_indices
+            .iter()
+            .position(|&line_end_byte| (line_end_byte as usize) > byte)
+            .unwrap_or(0);
+
+        let column = line
+            .checked_sub(1)
+            .and_then(|idx| line_end_indices.get(idx))
+            .map(|&prev_line_end| byte.saturating_sub(prev_line_end as usize))
+            .unwrap_or(byte);
+
+        Self::new(line, column, byte)
+    }
+
+    /// Shift the column by a given amount.
+    pub fn shift_column(self, column_move: usize) -> Self {
+        Self {
+            line: self.line,
+            character: self.character + column_move.saturating_sub(1),
+            byte_offset: 0,
+        }
+    }
+
+    /// Move to the next line.
+    pub fn move_to_next_line(mut self) -> Self {
+        self.line += 1;
+        self.character = 0;
+        self.byte_offset = 0;
+        self
+    }
+}
+
+impl From<tree_sitter::Point> for Position {
+    fn from(point: tree_sitter::Point) -> Self {
+        Self {
+            line: point.row,
+            character: point.column,
+            byte_offset: 0,
+        }
+    }
+}
+
+impl From<Position> for tree_sitter::Point {
+    fn from(val: Position) -> Self {
+        val.to_tree_sitter()
+    }
+}
+
+/// A range in a source file.
+///
+/// Ranges are stored as 0-indexed internally (tree-sitter compatible).
+#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Eq, Hash, Default)]
+#[serde(rename_all = "camelCase")]
+pub struct Range {
+    /// Start position of the range
+    start_position: Position,
+    /// End position of the range
+    end_position: Position,
+}
+
+impl Range {
+    /// Create a new range from start and end positions.
+    pub fn new(start_position: Position, end_position: Position) -> Self {
+        Self {
+            start_position,
+            end_position,
+        }
+    }
+
+    /// Create a range from a tree-sitter node.
+    pub fn for_tree_node(node: &tree_sitter::Node) -> Self {
+        let range = node.range();
+        Self {
+            start_position: Position {
+                line: range.start_point.row,
+                character: range.start_point.column,
+                byte_offset: range.start_byte,
+            },
+            end_position: Position {
+                line: range.end_point.row,
+                character: range.end_point.column,
+                byte_offset: range.end_byte,
+            },
+        }
+    }
+
+    /// Alias for for_tree_node (compatibility).
+    pub fn from_tree_sitter_node(node: &tree_sitter::Node) -> Self {
+        Self::for_tree_node(node)
+    }
+
+    /// Get the start position.
+    pub fn start_position(&self) -> Position {
+        self.start_position
+    }
+
+    /// Get the end position.
+    pub fn end_position(&self) -> Position {
+        self.end_position
+    }
+
+    /// Get the start position by reference.
+    pub fn get_start_position(&self) -> &Position {
+        &self.start_position
+    }
+
+    /// Get the end position by reference.
+    pub fn get_end_position(&self) -> &Position {
+        &self.end_position
+    }
+
+    /// Set the start position.
+    pub fn set_start_position(&mut self, position: Position) {
+        self.start_position = position;
+    }
+
+    /// Set the end position.
+    pub fn set_end_position(&mut self, position: Position) {
+        self.end_position = position;
+    }
+
+    /// Set the start byte offset.
+    pub fn set_start_byte(&mut self, byte: usize) {
+        self.start_position.set_byte_offset(byte);
+    }
+
+    /// Set the end byte offset.
+    pub fn set_end_byte(&mut self, byte: usize) {
+        self.end_position.set_byte_offset(byte);
+    }
+
+    /// Get the start byte offset.
+    pub fn start_byte(&self) -> usize {
+        self.start_position.byte_offset
+    }
+
+    /// Get the end byte offset.
+    pub fn end_byte(&self) -> usize {
+        self.end_position.byte_offset
+    }
+
+    /// Get the start line (0-indexed).
+    pub fn start_line(&self) -> usize {
+        self.start_position.line
+    }
+
+    /// Get the start line (1-indexed, for LSP/display).
+    pub fn start_line_1indexed(&self) -> usize {
+        self.start_position.line + 1
+    }
+
+    /// Get the end line (0-indexed).
+    pub fn end_line(&self) -> usize {
+        self.end_position.line
+    }
+
+    /// Get the end line (1-indexed, for LSP/display).
+    pub fn end_line_1indexed(&self) -> usize {
+        self.end_position.line + 1
+    }
+
+    /// Get the start column (0-indexed).
+    pub fn start_column(&self) -> usize {
+        self.start_position.character
+    }
+
+    /// Get the start column (1-indexed, for LSP/display).
+    pub fn start_column_1indexed(&self) -> usize {
+        self.start_position.character + 1
+    }
+
+    /// Get the end column (0-indexed).
+    pub fn end_column(&self) -> usize {
+        self.end_position.character
+    }
+
+    /// Get the end column (1-indexed, for LSP/display).
+    pub fn end_column_1indexed(&self) -> usize {
+        self.end_position.character + 1
+    }
+
+    /// Get the byte size of the range.
+    pub fn byte_size(&self) -> usize {
+        self.end_byte().saturating_sub(self.start_byte()) + 1
+    }
+
+    /// Get the number of bytes (alias).
+    pub fn len(&self) -> usize {
+        self.end_byte().saturating_sub(self.start_byte())
+    }
+
+    /// Check if the range is empty (zero bytes).
+    pub fn is_empty(&self) -> bool {
+        self.len() == 0
+    }
+
+    /// Get the line size (number of lines spanned, can be negative for inverted ranges).
+    pub fn line_size(&self) -> i64 {
+        self.end_line() as i64 - self.start_line() as i64
+    }
+
+    /// Check if this range contains another range (using line/column).
+    pub fn contains(&self, other: &Range) -> bool {
+        self.contains_check_line_column(other)
+    }
+
+    /// Check if this range contains another range (line-only check).
+    pub fn contains_check_line(&self, other: &Range) -> bool {
+        let start_position_check = self.start_line() <= other.start_line();
+        let end_position_check = self.end_line() >= other.end_line();
+        start_position_check && end_position_check
+    }
+
+    /// Check if this range contains another range (line and column check).
+    pub fn contains_check_line_column(&self, other: &Range) -> bool {
+        let start_position_check = self.start_line() < other.start_line()
+            || (self.start_line() == other.start_line()
+                && self.start_column() <= other.start_column());
+        let end_position_check = self.end_line() > other.end_line()
+            || (self.end_line() == other.end_line() && self.end_column() >= other.end_column());
+        start_position_check && end_position_check
+    }
+
+    /// Check if this range contains a position.
+    pub fn contains_position(&self, position: &Position) -> bool {
+        self.start_position().before_other(position) && self.end_position().after_other(position)
+    }
+
+    /// Check if this range contains a line.
+    pub fn contains_line(&self, line: usize) -> bool {
+        self.start_position().line() <= line && line <= self.end_position().line()
+    }
+
+    /// Check if this range intersects with another (line-based).
+    pub fn intersects_without_byte(&self, other: &Range) -> bool {
+        // Two ranges intersect if one starts before the other ends AND ends after the other starts
+        self.start_line() <= other.end_line() && self.end_line() >= other.start_line()
+    }
+
+    /// Convert to tree-sitter Range.
+    pub fn to_tree_sitter_range(&self) -> tree_sitter::Range {
+        tree_sitter::Range {
+            start_byte: self.start_position.byte_offset,
+            end_byte: self.end_position.byte_offset,
+            start_point: self.start_position.to_tree_sitter(),
+            end_point: self.end_position.to_tree_sitter(),
+        }
+    }
+
+    /// Create a range from byte offsets using line end indices.
+    pub fn from_byte_range(range: std::ops::Range<usize>, line_end_indices: &[u32]) -> Range {
+        let start = Position::from_byte(range.start, line_end_indices);
+        let end = Position::from_byte(range.end, line_end_indices);
+        Self::new(start, end)
+    }
+
+    /// Check equality based on line numbers only.
+    pub fn check_equality_without_byte(&self, other: &Range) -> bool {
+        self.start_line() == other.start_line() && self.end_line() == other.end_line()
+    }
+
+    /// Check equality based on line ranges.
+    pub fn equals_line_range(&self, other: &Range) -> bool {
+        self.start_line() == other.start_line() && self.end_line() == other.end_line()
+    }
+}