feat(session): auto-compact long transcripts via transformContext

pi's `transformContext` hook fired on every LLM call but returned
the transcript unchanged, so a long-running agent paid per-token
for its entire history on every turn. The README promised "sessions
persist + auto-compact" but the main loop had no compaction wired
in at all — only rotation at the long-cycle boundary.

`context-compactor` closes that gap. A pure, provider-free function
estimates tokens (~4 chars/token), short-circuits below a 120K
threshold, and on overflow keeps the last 20 messages verbatim
while collapsing older ones into a single "context summary" user
turn. The summary preserves the original task and the first twelve
assistant highlights so the agent has continuity without forwarding
every historical turn.

Compaction is non-destructive: the Agent's in-memory transcript
and the SQLite session store still hold the full history, feeding
UI timelines and audit. Only the LLM payload shrinks. Rotation
remains the long-cycle boundary (separate commit) — compaction is
what keeps each individual LLM call within budget.

src/main/agent-runtime/hook-handlers.ts

@@ -29,6 +29,7 @@ import {
   projectToolExecutionToWorkProduct,
   type ArtifactProjectorRunContext,
 } from "./artifact-projector";
+import { compactMessagesIfNeeded } from "./context-compactor";
 
 export interface ActiveRunContext {
   runId: string;
@@ -174,15 +175,16 @@ export function handleAfterToolCall(
 }
 
 /**
- * Context-transform hook. Placeholder pass-through; auto-compaction
- * wiring lives in session-management-service and plugs in here once
- * the compactor is available (tracked separately from the contract
- * wiring).
+ * Context-transform hook. Delegates to `compactMessagesIfNeeded`,
+ * which is a pure, provider-free deterministic compactor. Runs before
+ * every LLM call; short transcripts short-circuit on the token-budget
+ * check so the steady-state cost is negligible.
  */
 export function handleTransformContext(
   _deps: HookHandlerDeps,
   _ctx: ActiveRunContext,
-  _payload: TransformContextRpcPayload,
+  payload: TransformContextRpcPayload,
 ): TransformContextRpcResult {
-  return { changed: false, messages: null };
+  const outcome = compactMessagesIfNeeded(payload.messages);
+  return { changed: outcome.changed, messages: outcome.messages };
 }