gsd-pi 2.80.0-dev.c5f2443b3 → 2.80.0-dev.d4fc28e6b

package/src/resources/extensions/gsd/tests/compaction-snapshot.test.ts

@@ -29,7 +29,8 @@ test('buildSnapshot: renders memories, exec history, and active context', () =>
     memories: [
       { id: 'MEM001', category: 'gotcha', content: 'FTS5 needs Porter tokenizer', confidence: 0.9,
         source_unit_type: null, source_unit_id: null, created_at: '', updated_at: '',
-        superseded_by: null, hit_count: 0, scope: 'project', seq: 1, tags: [], structured_fields: null },
+        superseded_by: null, hit_count: 0, scope: 'project', seq: 1, tags: [], structured_fields: null,
+        last_hit_at: null },
     ],
     execHistory: [
       {
@@ -65,6 +66,7 @@ test('buildSnapshot: enforces the byte cap with a truncation marker', () => {
     seq: i,
     tags: [] as string[],
     structured_fields: null,
+    last_hit_at: null,
   }));
   const snap = buildSnapshot(
     { generatedAt: new Date(), memories: longMemories, execHistory: [] },
@@ -121,3 +123,14 @@ test('executeResume: reports friendly empty state when no snapshot exists', () =
     cleanup(base);
   }
 });
+
+test('executeResume: returns disabled error when context_mode.enabled=false', () => {
+  const base = freshBase();
+  try {
+    const result = executeResume({}, { baseDir: base, preferences: { context_mode: { enabled: false } } });
+    assert.equal(result.isError, true);
+    assert.equal((result.details as { error?: string }).error, 'context_mode_disabled');
+  } finally {
+    cleanup(base);
+  }
+});

package/src/resources/extensions/gsd/tests/complete-milestone-excerpt.test.ts

@@ -261,3 +261,34 @@ test("#4780 closer prompt: uses excerpts + lists on-demand slice SUMMARY paths",
     `closer prompt length ${prompt.length} should be < raw summary size ${rawSize} + 20KB headroom`,
   );
 });
+
+test("complete-milestone prompt caps repeated inlined context around 20k chars", async (t) => {
+  const base = createBase();
+  t.after(() => cleanup(base));
+  invalidateAllCaches();
+
+  writeRoadmap(base, makeRoadmap());
+  writeSummary(base, "S01", makeFatSummary("S01"));
+  writeSummary(base, "S02", makeFatSummary("S02"));
+  writeFileSync(
+    join(base, ".gsd", "milestones", "M001", "M001-CONTEXT.md"),
+    "# M001 Context\n\n" + "Large milestone context body. ".repeat(1200),
+  );
+  writeFileSync(
+    join(base, ".gsd", "KNOWLEDGE.md"),
+    "# Project Knowledge\n\n## Patterns\n\n### Test Milestone shared\n" + "Large scoped knowledge body. ".repeat(1200),
+  );
+
+  const prompt = await buildCompleteMilestonePrompt("M001", "Test Milestone", base);
+  const contextStart = prompt.indexOf("## Inlined Context (preloaded");
+  const contextEnd = prompt.indexOf("## Steps", contextStart);
+  assert.ok(contextStart >= 0, "prompt should include inlined context");
+  assert.ok(contextEnd > contextStart, "prompt should include steps after inlined context");
+
+  const inlinedContext = prompt.slice(contextStart, contextEnd);
+  assert.ok(
+    inlinedContext.length <= 21_000,
+    `inlined context ${inlinedContext.length} chars should stay near the 20k cap`,
+  );
+  assert.match(inlinedContext, /\[\.\.\.truncated \d+ sections\]/);
+});

package/src/resources/extensions/gsd/tests/complete-slice-composer.test.ts

@@ -113,8 +113,9 @@ test("#4782 phase 3: buildCompleteSlicePrompt composes roadmap → plan → task
     "task summaries precede slice-summary template",
   );
 
-  // Task body inlined
-  assert.match(prompt, /Task one did the thing/);
+  // Task summary excerpt is inlined; full narrative remains on-demand.
+  assert.match(prompt, /### Task Summary: T01 \(excerpt\)/);
+  assert.doesNotMatch(prompt, /Task one did the thing/);
 });
 
 test("#4782 phase 3: buildCompleteSlicePrompt handles missing task summaries gracefully", async (t) => {

package/src/resources/extensions/gsd/tests/context-budget.test.ts

@@ -4,7 +4,7 @@
  * No I/O, no extension context, no global state.
  */
 
-import { describe, it } from "node:test";
+import { describe, it, beforeEach } from "node:test";
 import assert from "node:assert/strict";
 
 import {
@@ -16,8 +16,17 @@ import {
   computeBudgets,
   truncateAtSectionBoundary,
   resolveExecutorContextWindow,
+  _resetEmpiricalCacheForTest,
 } from "../context-budget.js";
 
+// Reset the per-provider empirical chars-per-token cache before each test.
+// The hardcoded char-ratio assertions below assume the static fallback path
+// (3.5 / 4.0 chars/token) is used. Without this guard, a prior test that
+// warms tiktoken would populate the cache and silently break these tests.
+beforeEach(() => {
+  _resetEmpiricalCacheForTest();
+});
+
 import type { TokenProvider } from "../token-counter.js";
 
 // ─── Test helpers ─────────────────────────────────────────────────────────────

package/src/resources/extensions/gsd/tests/context-store.test.ts

@@ -1,3 +1,5 @@
+// GSD-2 + context-store.test.ts — Regression coverage for DB-backed context query helpers.
+
 import { describe, test, afterEach } from "node:test";
 import assert from "node:assert/strict";
 import {
@@ -362,7 +364,11 @@ describe("context-store: sub-5ms query timing", () => {
 
     assert.strictEqual(decisions.length, 50, `got ${decisions.length} decisions (expected 50)`);
     assert.strictEqual(requirements.length, 50, `got ${requirements.length} requirements (expected 50)`);
-    assert.ok(elapsed < 5, `query latency ${elapsed.toFixed(2)}ms should be < 5ms`);
+    const maxLatencyMs = process.env.NODE_V8_COVERAGE ? 15 : 5;
+    assert.ok(
+      elapsed < maxLatencyMs,
+      `query latency ${elapsed.toFixed(2)}ms should be < ${maxLatencyMs}ms`,
+    );
   });
 });
 

package/src/resources/extensions/gsd/tests/crash-handler-secondary.test.ts

@@ -232,4 +232,59 @@ describe('emitCrashRecoveredUnitEnd (#3348)', () => {
       rmSync(base, { recursive: true, force: true });
     }
   });
+
+  test('emitOpenUnitEndForUnit closes the latest open start with error context', async () => {
+    const base = makeTmpBase();
+    try {
+      const { emitJournalEvent, queryJournal } = await import('../journal.ts');
+      const { emitOpenUnitEndForUnit } = await import('../crash-recovery.ts');
+
+      const firstFlowId = randomUUID();
+      const secondFlowId = randomUUID();
+      emitJournalEvent(base, {
+        ts: new Date().toISOString(),
+        flowId: firstFlowId,
+        seq: 1,
+        eventType: 'unit-start',
+        data: { unitType: 'execute-task', unitId: 'M008/S04/T02' },
+      });
+      emitJournalEvent(base, {
+        ts: new Date().toISOString(),
+        flowId: firstFlowId,
+        seq: 2,
+        eventType: 'unit-end',
+        data: { unitType: 'execute-task', unitId: 'M008/S04/T02', status: 'completed' },
+        causedBy: { flowId: firstFlowId, seq: 1 },
+      });
+      emitJournalEvent(base, {
+        ts: new Date().toISOString(),
+        flowId: secondFlowId,
+        seq: 3,
+        eventType: 'unit-start',
+        data: { unitType: 'execute-task', unitId: 'M008/S04/T02' },
+      });
+
+      const emitted = emitOpenUnitEndForUnit(
+        base,
+        'execute-task',
+        'M008/S04/T02',
+        'cancelled',
+        { message: 'runUnitPhase exploded', category: 'unit-exception', isTransient: false },
+      );
+
+      assert.equal(emitted, true, 'open unit should be closed');
+      const ends = queryJournal(base).filter((e) => e.eventType === 'unit-end');
+      assert.equal(ends.length, 2, 'should preserve existing end and add one new end');
+      const newEnd = ends.find((e) => e.causedBy?.flowId === secondFlowId);
+      assert.ok(newEnd, 'new end should close the latest open start');
+      assert.equal(newEnd!.data?.status, 'cancelled');
+      assert.deepEqual(newEnd!.data?.errorContext, {
+        message: 'runUnitPhase exploded',
+        category: 'unit-exception',
+        isTransient: false,
+      });
+    } finally {
+      rmSync(base, { recursive: true, force: true });
+    }
+  });
 });

package/src/resources/extensions/gsd/tests/crash-recovery-via-db.test.ts

@@ -30,6 +30,7 @@ import {
   isLockProcessAlive,
 } from "../crash-recovery.ts";
 import { normalizeRealPath } from "../paths.ts";
+import { writeUnitRuntimeRecord } from "../unit-runtime.ts";
 
 function makeBase(): string {
   const base = mkdtempSync(join(tmpdir(), "gsd-crash-recovery-"));
@@ -102,6 +103,27 @@ test("readCrashLock synthesizes LockData from a stale dead worker (no dispatches
   assert.ok(lock!.startedAt, "startedAt populated from workers.started_at");
 });
 
+test("readCrashLock falls back to latest in-flight runtime record when dispatch claim is missing", (t) => {
+  const base = makeBase();
+  t.after(() => cleanup(base));
+  openDatabase(join(base, ".gsd", "gsd.db"));
+  const projectRoot = normalizeRealPath(base);
+  const workerId = registerAutoWorker({ projectRootRealpath: projectRoot });
+  writeUnitRuntimeRecord(base, "execute-task", "M008/S04/T02", 1778069087937, {
+    phase: "dispatched",
+    lastProgressAt: 1778069087937,
+    lastProgressKind: "dispatch",
+  });
+  setWorkerPid(workerId, 99999);
+  expireWorker(workerId);
+
+  const lock = readCrashLock(base);
+  assert.ok(lock, "stale worker surfaced as a crash lock");
+  assert.equal(lock!.unitType, "execute-task");
+  assert.equal(lock!.unitId, "M008/S04/T02");
+  assert.equal(lock!.unitStartedAt, new Date(1778069087937).toISOString());
+});
+
 test("readCrashLock includes the most recent dispatch as unitType/unitId", (t) => {
   const base = makeBase();
   t.after(() => cleanup(base));

package/src/resources/extensions/gsd/tests/custom-engine-loop-integration.test.ts

@@ -192,7 +192,11 @@ function makeMockDeps(overrides?: Partial<LoopDeps>): LoopDeps & { callLog: stri
     resolveMilestoneFile: () => null,
     reconcileMergeState: () => "clean",
     preflightCleanRoot: () => ({ stashPushed: false, summary: "" }),
-    postflightPopStash: () => {},
+    postflightPopStash: () => ({
+      restored: true,
+      needsManualRecovery: false,
+      message: "restored",
+    }),
     getLedger: () => null,
     getProjectTotals: () => ({ cost: 0 }),
     formatCost: (c: number) => `$${c.toFixed(2)}`,
@@ -372,7 +376,56 @@ describe("Custom engine loop integration", () => {
     assert.ok(stopEntry?.includes("Workflow complete"), "Should stop with 'Workflow complete'");
   });
 
-  it("finishes custom-engine skip turns and clears current turn state", async () => {
+  it("finalizes custom-engine complete turns and clears current turn state", async () => {
+    _resetPendingResolve();
+
+    const runDir = makeTmpDir();
+    const graph = makeGraph([
+      makeStep({ id: "step-a", status: "complete" }),
+    ], "already-done");
+    writeGraph(runDir, graph);
+    writeDefinition(runDir, graph.steps, "already-done");
+
+    const ctx = makeMockCtx();
+    const pi = makeMockPi();
+    const s = makeLoopSession({
+      activeEngineId: "custom",
+      activeRunDir: runDir,
+      basePath: runDir,
+    });
+    const turnResults: Array<{ status: string; failureClass: string; error?: string }> = [];
+    const deps = makeMockDeps({
+      stopAuto: async (_ctx, _pi, reason) => {
+        deps.callLog.push(`stopAuto:${reason ?? "no-reason"}`);
+        s.active = false;
+      },
+      uokObserver: {
+        onTurnStart: () => {},
+        onTurnResult: (result) => {
+          deps.callLog.push(`turnResult:${result.status}`);
+          turnResults.push({
+            status: result.status,
+            failureClass: result.failureClass,
+            error: result.error,
+          });
+        },
+        onPhaseResult: () => {},
+      },
+    });
+
+    await autoLoop(ctx, pi, s, deps);
+
+    assert.deepEqual(turnResults, [{ status: "completed", failureClass: "none", error: undefined }]);
+    assert.ok(
+      deps.callLog.indexOf("turnResult:completed") < deps.callLog.indexOf("stopAuto:Workflow complete"),
+      `turn should finalize before stopAuto; log=${deps.callLog.join(",")}`,
+    );
+    assert.equal(s.currentTraceId, null);
+    assert.equal(s.currentTurnId, null);
+    assert.equal(pi.calls.length, 0, "complete workflow should not dispatch work");
+  });
+
+  it("stops blocked custom workflows and clears current turn state", async () => {
     _resetPendingResolve();
 
     const runDir = makeTmpDir();
@@ -390,13 +443,65 @@ describe("Custom engine loop integration", () => {
       activeRunDir: runDir,
       basePath: runDir,
     });
-    const turnResults: Array<{ status: string }> = [];
+    const turnResults: Array<{ status: string; failureClass: string; error?: string }> = [];
     const deps = makeMockDeps({
+      stopAuto: async (_ctx, _pi, reason) => {
+        deps.callLog.push(`stopAuto:${reason ?? "no-reason"}`);
+        s.active = false;
+      },
+      uokObserver: {
+        onTurnStart: () => {},
+        onTurnResult: (result) => {
+          turnResults.push({
+            status: result.status,
+            failureClass: result.failureClass,
+            error: result.error,
+          });
+        },
+        onPhaseResult: () => {},
+      },
+    });
+
+    await autoLoop(ctx, pi, s, deps);
+
+    assert.equal(turnResults.length, 1);
+    assert.equal(turnResults[0].status, "stopped");
+    assert.equal(turnResults[0].failureClass, "manual-attention");
+    assert.match(turnResults[0].error ?? "", /custom-engine-dispatch-stop/);
+    assert.equal(s.currentTraceId, null);
+    assert.equal(s.currentTurnId, null);
+    assert.equal(pi.calls.length, 0, "blocked workflow should not dispatch a custom step");
+    assert.match(
+      deps.callLog.find((e: string) => e.startsWith("stopAuto:")) ?? "",
+      /Workflow blocked: no pending steps are ready/,
+    );
+  });
+
+  it("finalizes the active turn when the session lock is lost", async () => {
+    _resetPendingResolve();
+
+    const ctx = makeMockCtx();
+    const pi = makeMockPi();
+    const s = makeLoopSession();
+    const turnResults: Array<{ status: string; failureClass: string; error?: string }> = [];
+    const deps = makeMockDeps({
+      validateSessionLock: () => ({
+        valid: false,
+        failureReason: "pid-mismatch",
+        expectedPid: 111,
+        existingPid: 222,
+      } as SessionLockStatus),
+      handleLostSessionLock: () => {
+        deps.callLog.push("handleLostSessionLock");
+      },
       uokObserver: {
         onTurnStart: () => {},
         onTurnResult: (result) => {
-          turnResults.push({ status: result.status });
-          s.active = false;
+          turnResults.push({
+            status: result.status,
+            failureClass: result.failureClass,
+            error: result.error,
+          });
         },
         onPhaseResult: () => {},
       },
@@ -404,10 +509,15 @@ describe("Custom engine loop integration", () => {
 
     await autoLoop(ctx, pi, s, deps);
 
-    assert.deepEqual(turnResults, [{ status: "skipped" }]);
+    assert.deepEqual(turnResults, [{
+      status: "stopped",
+      failureClass: "manual-attention",
+      error: "session-lock-lost",
+    }]);
     assert.equal(s.currentTraceId, null);
     assert.equal(s.currentTurnId, null);
-    assert.equal(pi.calls.length, 0, "skip should not dispatch a custom step");
+    assert.equal(pi.calls.length, 0, "lost session lock must not dispatch work");
+    assert.ok(deps.callLog.includes("handleLostSessionLock"));
   });
 
   it("does not call runPreDispatch or runFinalize on the custom path", async () => {

package/src/resources/extensions/gsd/tests/custom-workflow-engine.test.ts

@@ -193,7 +193,7 @@ describe("CustomWorkflowEngine.resolveDispatch", () => {
     }
   });
 
-  it("returns skip when pending steps are blocked by incomplete dependencies", async () => {
+  it("returns stop with dependency details when pending steps are blocked", async () => {
     const { engine } = setupEngine([
       makeStep({ id: "a", dependsOn: ["b"] }),
       makeStep({ id: "b", dependsOn: ["a"] }),
@@ -202,7 +202,45 @@ describe("CustomWorkflowEngine.resolveDispatch", () => {
     const state = await engine.deriveState("/unused");
     const dispatch = await engine.resolveDispatch(state, { basePath: "/unused" });
 
-    assert.deepEqual(dispatch, { action: "skip" });
+    assert.equal(dispatch.action, "stop");
+    if (dispatch.action === "stop") {
+      assert.equal(dispatch.level, "error");
+      assert.match(dispatch.reason, /Workflow blocked/);
+      assert.match(dispatch.reason, /a waiting on b \(pending\)/);
+      assert.match(dispatch.reason, /b waiting on a \(pending\)/);
+    }
+  });
+
+  it("reports missing dependencies when no pending step can run", async () => {
+    const { engine } = setupEngine([
+      makeStep({ id: "a", dependsOn: ["missing-step"] }),
+    ]);
+
+    const state = await engine.deriveState("/unused");
+    const dispatch = await engine.resolveDispatch(state, { basePath: "/unused" });
+
+    assert.equal(dispatch.action, "stop");
+    if (dispatch.action === "stop") {
+      assert.equal(dispatch.level, "error");
+      assert.match(dispatch.reason, /a waiting on missing-step \(missing\)/);
+    }
+  });
+
+  it("does not report expanded dependencies as blockers", async () => {
+    const { engine } = setupEngine([
+      makeStep({ id: "iter", status: "expanded" }),
+      makeStep({ id: "after", dependsOn: ["iter", "missing-step"] }),
+    ]);
+
+    const state = await engine.deriveState("/unused");
+    const dispatch = await engine.resolveDispatch(state, { basePath: "/unused" });
+
+    assert.equal(dispatch.action, "stop");
+    if (dispatch.action === "stop") {
+      assert.equal(dispatch.level, "error");
+      assert.match(dispatch.reason, /after waiting on missing-step \(missing\)/);
+      assert.doesNotMatch(dispatch.reason, /iter \(expanded\)/);
+    }
   });
 
   it("respects dependency ordering", async () => {