feat(02-01): create canonical key derivation and skill source resolution

- deriveCanonicalSkillKey handles all 5 source types deterministically
- normalizeSkillSlug with lowercase, hyphen-delimited, edge case handling
- hashSkillValue produces 10-char SHA-256 hex digests
- parseSkillImportSource resolves GitHub, skills.sh, URL, local path inputs
This commit is contained in:
ZacharyZhang-NY committed 2026-03-26 00:27:39 -04:00
1 parent 77480374fd
commit a8db40180c
1 file changed
+194
+194
View File
@@ -0,0 +1,194 @@
// ---------------------------------------------------------------------------
// Canonical skill key derivation and skill import source resolution
// ---------------------------------------------------------------------------
import { createHash } from "node:crypto";
import { resolve } from "node:path";
import type { SkillSourceType } from "./skill-types.js";
// ---------------------------------------------------------------------------
// Slug normalization helpers
// ---------------------------------------------------------------------------
const SLUG_DELIM_RE = /[^a-z0-9]+/g;
const SLUG_TRIM_RE = /^-+|-+$/g;
/**
* Normalize a value into a URL-safe slug.
* Returns null if the value is falsy or results in an empty string.
*/
export function normalizeSkillSlug(
value: string | undefined | null,
): string | null {
if (!value) return null;
const trimmed = value.trim();
if (!trimmed) return null;
const normalized = trimmed
.toLowerCase()
.replace(SLUG_DELIM_RE, "-")
.replace(SLUG_TRIM_RE, "");
return normalized.length > 0 ? normalized : null;
}
// ---------------------------------------------------------------------------
// Hashing
// ---------------------------------------------------------------------------
/**
* Compute the first 10 characters of a SHA-256 hex digest.
* Used for hash segments in canonical keys.
*/
export function hashSkillValue(value: string): string {
return createHash("sha256").update(value).digest("hex").slice(0, 10);
}
// ---------------------------------------------------------------------------
// Canonical key derivation
// ---------------------------------------------------------------------------
function asString(value: unknown): string | null {
return typeof value === "string" && value.length > 0 ? value : null;
}
function isPlainRecord(value: unknown): value is Record<string, unknown> {
return typeof value === "object" && value !== null && !Array.isArray(value);
}
/**
* Derive a deterministic canonical key for a skill.
*
* Key formats by source type:
* - GitHub/skills_sh with owner+repo: `{owner}/{repo}/{slug}`
* - URL with sourceLocator: `url/{normalizedHost}/{hash10}/{slug}`
* - local_path with sourceLocator: `local/{hash10 of resolved path}/{slug}`
* - Fallback (managed or unknown): `company/{companyId}/{slug}`
*/
export function deriveCanonicalSkillKey(
companyId: string,
input: {
slug: string;
sourceType: SkillSourceType;
sourceLocator: string | null;
metadata: Record<string, unknown> | null;
},
): string {
const slug = normalizeSkillSlug(input.slug) ?? "skill";
const metadata = isPlainRecord(input.metadata) ? input.metadata : null;
const owner = normalizeSkillSlug(asString(metadata?.owner));
const repo = normalizeSkillSlug(asString(metadata?.repo));
// GitHub or skills.sh with owner/repo metadata
if (
(input.sourceType === "github" || input.sourceType === "skills_sh") &&
owner &&
repo
) {
return `${owner}/${repo}/${slug}`;
}
// URL source with locator
if (input.sourceType === "url") {
const locator = asString(input.sourceLocator);
if (locator) {
try {
const url = new URL(locator);
const host = normalizeSkillSlug(url.host) ?? "url";
return `url/${host}/${hashSkillValue(locator)}/${slug}`;
} catch {
return `url/unknown/${hashSkillValue(locator)}/${slug}`;
}
}
}
// Local path source with locator
if (input.sourceType === "local_path") {
const locator = asString(input.sourceLocator);
if (locator) {
return `local/${hashSkillValue(resolve(locator))}/${slug}`;
}
}
// Fallback: company-scoped key
return `company/${companyId}/${slug}`;
}
// ---------------------------------------------------------------------------
// Skill import source resolution
// ---------------------------------------------------------------------------
// Three-part key: org/repo/skill
const THREE_PART_KEY_RE =
/^[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+$/;
// Two-part key: org/repo
const TWO_PART_KEY_RE = /^[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+$/;
// skills.sh URL pattern
const SKILLS_SH_RE =
/^https?:\/\/(?:www\.)?skills\.sh\/([A-Za-z0-9_.-]+)\/([A-Za-z0-9_.-]+)(?:\/([A-Za-z0-9_.-]+))?(?:[?#].*)?$/i;
export interface ParsedSkillImportSource {
resolvedSource: string;
requestedSkillSlug: string | null;
originalSkillsShUrl: string | null;
}
/**
* Parse a raw skill import source string and resolve it to a normalized form.
*
* Supported input formats:
* - `owner/repo/skill` -> GitHub URL with skills.sh origin tracking
* - `owner/repo` -> GitHub URL
* - `https://skills.sh/owner/repo/skill` -> GitHub URL with skills.sh origin
* - `https://github.com/owner/repo` -> pass through
* - `/local/path` -> pass through
*/
export function parseSkillImportSource(
rawInput: string,
): ParsedSkillImportSource {
const trimmed = rawInput.trim();
// Three-part key (org/repo/skill) -- originates from skills.sh registry
if (
!/^https?:\/\//i.test(trimmed) &&
THREE_PART_KEY_RE.test(trimmed)
) {
const [owner, repo, skillSlugRaw] = trimmed.split("/");
return {
resolvedSource: `https://github.com/${owner}/${repo}`,
requestedSkillSlug: normalizeSkillSlug(skillSlugRaw),
originalSkillsShUrl: `https://skills.sh/${owner}/${repo}/${skillSlugRaw}`,
};
}
// Two-part key (org/repo) -- resolve to GitHub
if (
!/^https?:\/\//i.test(trimmed) &&
TWO_PART_KEY_RE.test(trimmed)
) {
return {
resolvedSource: `https://github.com/${trimmed}`,
requestedSkillSlug: null,
originalSkillsShUrl: null,
};
}
// skills.sh URL -> resolve to GitHub
const skillsShMatch = trimmed.match(SKILLS_SH_RE);
if (skillsShMatch) {
const [, owner, repo, skillSlugRaw] = skillsShMatch;
return {
resolvedSource: `https://github.com/${owner}/${repo}`,
requestedSkillSlug: skillSlugRaw
? normalizeSkillSlug(skillSlugRaw)
: null,
originalSkillsShUrl: trimmed,
};
}
// Everything else -- pass through as resolvedSource
return {
resolvedSource: trimmed,
requestedSkillSlug: null,
originalSkillsShUrl: null,
};
}