gsd-pi 2.47.0-dev.8cfe772 → 2.47.0-dev.f2e721d

package/src/resources/extensions/gsd/tests/forensics-journal.test.ts

@@ -0,0 +1,162 @@
+import { describe, it } from "node:test";
+import assert from "node:assert/strict";
+import { readFileSync } from "node:fs";
+import { join, dirname } from "node:path";
+import { fileURLToPath } from "node:url";
+
+const __dirname = dirname(fileURLToPath(import.meta.url));
+const gsdDir = join(__dirname, "..");
+
+describe("forensics journal & activity log awareness", () => {
+  const forensicsSrc = readFileSync(join(gsdDir, "forensics.ts"), "utf-8");
+  const promptSrc = readFileSync(join(gsdDir, "prompts", "forensics.md"), "utf-8");
+
+  it("scanJournalForForensics reads journal files directly (no full queryJournal load)", () => {
+    // Must NOT use queryJournal which loads ALL entries into memory
+    assert.ok(
+      !forensicsSrc.includes('queryJournal('),
+      "forensics.ts must NOT call queryJournal() which loads all entries at once",
+    );
+    // Must have its own journal scanning with file-level limits
+    assert.ok(
+      forensicsSrc.includes("scanJournalForForensics"),
+      "forensics.ts must have scanJournalForForensics function",
+    );
+  });
+
+  it("journal scanning limits files parsed to avoid memory bloat", () => {
+    assert.ok(
+      forensicsSrc.includes("MAX_JOURNAL_RECENT_FILES"),
+      "must have MAX_JOURNAL_RECENT_FILES constant to limit parsed files",
+    );
+    assert.ok(
+      forensicsSrc.includes("MAX_JOURNAL_RECENT_EVENTS"),
+      "must have MAX_JOURNAL_RECENT_EVENTS constant to limit events extracted",
+    );
+  });
+
+  it("older journal files are line-counted without full JSON parse", () => {
+    assert.ok(
+      forensicsSrc.includes("olderEntryCount") || forensicsSrc.includes("olderFiles"),
+      "must handle older files separately from recent files",
+    );
+  });
+
+  it("ForensicReport includes journalSummary field", () => {
+    assert.ok(
+      forensicsSrc.includes("journalSummary"),
+      "ForensicReport must include journalSummary field",
+    );
+  });
+
+  it("ForensicReport includes activityLogMeta field", () => {
+    assert.ok(
+      forensicsSrc.includes("activityLogMeta"),
+      "ForensicReport must include activityLogMeta field",
+    );
+  });
+
+  it("buildForensicReport calls scanJournalForForensics", () => {
+    assert.ok(
+      forensicsSrc.includes("scanJournalForForensics"),
+      "buildForensicReport must call scanJournalForForensics",
+    );
+  });
+
+  it("buildForensicReport calls gatherActivityLogMeta", () => {
+    assert.ok(
+      forensicsSrc.includes("gatherActivityLogMeta"),
+      "buildForensicReport must call gatherActivityLogMeta",
+    );
+  });
+
+  it("forensics detects journal-based anomalies", () => {
+    assert.ok(
+      forensicsSrc.includes("detectJournalAnomalies"),
+      "forensics.ts must have detectJournalAnomalies function",
+    );
+    // Check for specific journal anomaly types
+    assert.ok(forensicsSrc.includes('"journal-stuck"'), "must detect journal-stuck anomalies");
+    assert.ok(forensicsSrc.includes('"journal-guard-block"'), "must detect journal-guard-block anomalies");
+    assert.ok(forensicsSrc.includes('"journal-rapid-iterations"'), "must detect journal-rapid-iterations anomalies");
+    assert.ok(forensicsSrc.includes('"journal-worktree-failure"'), "must detect journal-worktree-failure anomalies");
+  });
+
+  it("formatReportForPrompt includes journal summary section", () => {
+    assert.ok(
+      forensicsSrc.includes("Journal Summary"),
+      "prompt formatter must include a Journal Summary section",
+    );
+  });
+
+  it("formatReportForPrompt includes activity log overview section", () => {
+    assert.ok(
+      forensicsSrc.includes("Activity Log Overview"),
+      "prompt formatter must include an Activity Log Overview section",
+    );
+  });
+
+  it("activity log scanning uses tail-read with byte cap (not full file load)", () => {
+    // scanActivityLogs uses nativeParseJsonlTail + MAX_JSONL_BYTES for efficient reading
+    assert.ok(
+      forensicsSrc.includes("nativeParseJsonlTail"),
+      "activity log scanning must use nativeParseJsonlTail for tail-reading",
+    );
+    assert.ok(
+      forensicsSrc.includes("MAX_JSONL_BYTES"),
+      "activity log scanning must respect MAX_JSONL_BYTES cap",
+    );
+    // Only reads last 5 files
+    assert.ok(
+      forensicsSrc.includes("slice(-5)"),
+      "activity log scanning must limit to last 5 files",
+    );
+  });
+
+  it("activity log entries are distilled through extractTrace, not sent raw", () => {
+    assert.ok(
+      forensicsSrc.includes("extractTrace("),
+      "activity log entries must be distilled through extractTrace before reporting",
+    );
+  });
+
+  it("prompt output is hard-capped at 30KB", () => {
+    assert.ok(
+      forensicsSrc.includes("MAX_BYTES") && forensicsSrc.includes("30 * 1024"),
+      "formatReportForPrompt must have a 30KB hard cap",
+    );
+    assert.ok(
+      forensicsSrc.includes("truncated at 30KB"),
+      "prompt must show truncation message when capped",
+    );
+  });
+
+  it("forensics prompt documents journal format", () => {
+    assert.ok(
+      promptSrc.includes("### Journal Format"),
+      "forensics.md must document the journal format",
+    );
+    assert.ok(
+      promptSrc.includes("flowId"),
+      "forensics.md must reference flowId concept",
+    );
+    assert.ok(
+      promptSrc.includes("causedBy"),
+      "forensics.md must reference causedBy for causal chains",
+    );
+  });
+
+  it("forensics prompt includes journal directory in runtime path reference", () => {
+    assert.ok(
+      promptSrc.includes("journal/"),
+      "forensics.md runtime path reference must include journal/",
+    );
+  });
+
+  it("investigation protocol references journal data", () => {
+    assert.ok(
+      promptSrc.includes("journal timeline") || promptSrc.includes("journal events"),
+      "investigation protocol must reference journal data for tracing",
+    );
+  });
+});

package/src/resources/extensions/gsd/tests/stale-milestone-id-reservation.test.ts

@@ -0,0 +1,79 @@
+/**
+ * Regression test for #2488: Stale milestone ID reservations inflate next ID
+ * after cancelled /gsd sessions.
+ *
+ * The module-level `reservedMilestoneIds` Set persists across /gsd invocations
+ * within the same Node process. Without clearReservedMilestoneIds() at session
+ * start, each cancelled session permanently bumps the counter by 1.
+ */
+import { describe, test, beforeEach } from "node:test";
+import assert from "node:assert/strict";
+
+import {
+  nextMilestoneId,
+  reserveMilestoneId,
+  getReservedMilestoneIds,
+  clearReservedMilestoneIds,
+} from "../milestone-ids.ts";
+
+describe("stale milestone ID reservation cleanup (#2488)", () => {
+  beforeEach(() => {
+    clearReservedMilestoneIds();
+  });
+
+  test("without cleanup, cancelled sessions inflate the next ID", () => {
+    const diskIds = ["M001", "M002", "M003"];
+
+    // Session 1: user starts /gsd, ID is previewed and reserved, then cancelled
+    const allIds1 = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const preview1 = nextMilestoneId(allIds1);
+    reserveMilestoneId(preview1);
+    assert.equal(preview1, "M004");
+
+    // Session 2: user starts /gsd again — stale reservation still in Set
+    // WITHOUT clearing, the next ID skips M004 (reserved) and goes to M005
+    const allIds2 = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const preview2 = nextMilestoneId(allIds2);
+    assert.equal(preview2, "M005", "without cleanup, ID inflates to M005");
+  });
+
+  test("with cleanup at session start, next ID is correct", () => {
+    const diskIds = ["M001", "M002", "M003"];
+
+    // Session 1: user starts /gsd, ID is previewed and reserved, then cancelled
+    const allIds1 = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const preview1 = nextMilestoneId(allIds1);
+    reserveMilestoneId(preview1);
+    assert.equal(preview1, "M004");
+
+    // Session 2: clear stale reservations first (the fix)
+    clearReservedMilestoneIds();
+
+    // Now the next ID correctly returns M004 again
+    const allIds2 = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const preview2 = nextMilestoneId(allIds2);
+    assert.equal(preview2, "M004", "after cleanup, ID is correctly M004");
+  });
+
+  test("multiple cancelled sessions compound the inflation without cleanup", () => {
+    const diskIds = ["M001", "M002", "M003"];
+
+    // 3 cancelled sessions without cleanup
+    for (let i = 0; i < 3; i++) {
+      const allIds = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+      const preview = nextMilestoneId(allIds);
+      reserveMilestoneId(preview);
+    }
+
+    // Without cleanup, we're now at M007 instead of M004
+    const allIds = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const next = nextMilestoneId(allIds);
+    assert.equal(next, "M007", "3 cancelled sessions inflate ID by 3");
+
+    // With cleanup, we're back to M004
+    clearReservedMilestoneIds();
+    const allIdsClean = [...new Set([...diskIds, ...getReservedMilestoneIds()])];
+    const nextClean = nextMilestoneId(allIdsClean);
+    assert.equal(nextClean, "M004", "cleanup restores correct next ID");
+  });
+});