test: harden beta release gates

2026-05-06 06:30:42 +00:00 · 2026-04-15 19:28:18 +01:00
parent 41699cfc2d
commit 4efd3c3d74
7 changed files with 95 additions and 46 deletions
--- a/extensions/memory-core/src/memory/index.test.ts
+++ b/extensions/memory-core/src/memory/index.test.ts
@@ -376,6 +376,9 @@ describe("memory index", () => {
    const manager = requireManager(result);
    managersForCleanup.add(manager);
    resetManagerForTest(manager);
+    if (!manager.status().fts?.available) {
+      return;
+    }

    await fs.writeFile(
      path.join(memoryDir, "2026-01-12.md"),
@@ -411,6 +414,9 @@ describe("memory index", () => {
      const manager = requireManager(result);
      managersForCleanup.add(manager);
      resetManagerForTest(manager);
+      if (!manager.status().fts?.available) {
+        return;
+      }

      const memoryPath = path.join(workspaceDir, "MEMORY.md");
      await fs.writeFile(memoryPath, "Project Nebula stale codename: ORBIT-9.\n", "utf8");
@@ -478,6 +484,9 @@ describe("memory index", () => {
      const manager = requireManager(result);
      managersForCleanup.add(manager);
      resetManagerForTest(manager);
+      if (!manager.status().fts?.available) {
+        return;
+      }

      const sessionsDir = resolveSessionTranscriptsDirForAgent("main");
      await fs.mkdir(sessionsDir, { recursive: true });
--- a/extensions/memory-core/src/memory/manager-search.test.ts
+++ b/extensions/memory-core/src/memory/manager-search.test.ts
@@ -9,9 +9,26 @@ import { searchKeyword } from "./manager-search.js";
 describe("searchKeyword trigram fallback", () => {
  const { DatabaseSync } = requireNodeSqlite();

+  function supportsTrigramFts(): boolean {
+    const db = new DatabaseSync(":memory:");
+    try {
+      const result = ensureMemoryIndexSchema({
+        db,
+        embeddingCacheTable: "embedding_cache",
+        cacheEnabled: false,
+        ftsTable: "chunks_fts",
+        ftsEnabled: true,
+        ftsTokenizer: "trigram",
+      });
+      return result.ftsAvailable;
+    } finally {
+      db.close();
+    }
+  }
+
  function createTrigramDb() {
    const db = new DatabaseSync(":memory:");
-    ensureMemoryIndexSchema({
+    const result = ensureMemoryIndexSchema({
      db,
      embeddingCacheTable: "embedding_cache",
      cacheEnabled: false,
@@ -19,6 +36,10 @@ describe("searchKeyword trigram fallback", () => {
      ftsEnabled: true,
      ftsTokenizer: "trigram",
    });
+    if (!result.ftsAvailable) {
+      db.close();
+      throw new Error(`FTS5 trigram unavailable: ${result.ftsError ?? "unknown error"}`);
+    }
    return db;
  }

@@ -53,7 +74,9 @@ describe("searchKeyword trigram fallback", () => {
    }
  }

-  it("finds short Chinese queries with substring fallback", async () => {
+  const itWithTrigramFts = supportsTrigramFts() ? it : it.skip;
+
+  itWithTrigramFts("finds short Chinese queries with substring fallback", async () => {
    const results = await runSearch({
      rows: [{ id: "1", path: "memory/zh.md", text: "今天玩成语接龙游戏" }],
      query: "成语",
@@ -62,7 +85,7 @@ describe("searchKeyword trigram fallback", () => {
    expect(results[0]?.textScore).toBe(1);
  });

-  it("finds short Japanese and Korean queries with substring fallback", async () => {
+  itWithTrigramFts("finds short Japanese and Korean queries with substring fallback", async () => {
    const japaneseResults = await runSearch({
      rows: [{ id: "jp", path: "memory/jp.md", text: "今日はしりとり大会" }],
      query: "しり とり",
@@ -76,19 +99,22 @@ describe("searchKeyword trigram fallback", () => {
    expect(koreanResults.map((row) => row.id)).toEqual(["ko"]);
  });

-  it("keeps MATCH semantics for long trigram terms while requiring short CJK substrings", async () => {
-    const results = await runSearch({
-      rows: [
-        { id: "match", path: "memory/good.md", text: "今天玩成语接龙游戏" },
-        { id: "partial", path: "memory/partial.md", text: "今天玩成语接龙" },
-      ],
-      query: "成语接龙 游戏",
-    });
-    expect(results.map((row) => row.id)).toEqual(["match"]);
-    expect(results[0]?.textScore).toBeGreaterThan(0);
-  });
+  itWithTrigramFts(
+    "keeps MATCH semantics for long trigram terms while requiring short CJK substrings",
+    async () => {
+      const results = await runSearch({
+        rows: [
+          { id: "match", path: "memory/good.md", text: "今天玩成语接龙游戏" },
+          { id: "partial", path: "memory/partial.md", text: "今天玩成语接龙" },
+        ],
+        query: "成语接龙 游戏",
+      });
+      expect(results.map((row) => row.id)).toEqual(["match"]);
+      expect(results[0]?.textScore).toBeGreaterThan(0);
+    },
+  );

-  it("applies fallback lexical boosts without exceeding bounded scores", async () => {
+  itWithTrigramFts("applies fallback lexical boosts without exceeding bounded scores", async () => {
    const results = await runSearch({
      rows: [
        {
@@ -133,7 +159,7 @@ describe("searchKeyword trigram fallback", () => {
    expect(boostedById.get("weak")?.score).toBeLessThanOrEqual(1);
  });

-  it("does not overweight repeated query tokens in fallback scoring", async () => {
+  itWithTrigramFts("does not overweight repeated query tokens in fallback scoring", async () => {
    const unique = await runSearch({
      rows: [{ id: "1", path: "memory/project.md", text: "Project memory context." }],
      query: "project memory context",
--- a/src/agents/pi-embedded-runner/run.overflow-compaction.harness.ts
+++ b/src/agents/pi-embedded-runner/run.overflow-compaction.harness.ts
@@ -72,6 +72,7 @@ export const mockedRunEmbeddedAttempt =
  vi.fn<(params: unknown) => Promise<EmbeddedRunAttemptResult>>();
 export const mockedRunContextEngineMaintenance = vi.fn(async () => undefined);
 export const mockedSessionLikelyHasOversizedToolResults = vi.fn(() => false);
+export const mockedResolveLiveToolResultMaxChars = vi.fn(() => 32_000);
 type MockTruncateOversizedToolResultsResult = {
  truncated: boolean;
  truncatedCount: number;
@@ -228,6 +229,8 @@ export function resetRunOverflowCompactionHarnessMocks(): void {
  mockedRunContextEngineMaintenance.mockResolvedValue(undefined);
  mockedSessionLikelyHasOversizedToolResults.mockReset();
  mockedSessionLikelyHasOversizedToolResults.mockReturnValue(false);
+  mockedResolveLiveToolResultMaxChars.mockReset();
+  mockedResolveLiveToolResultMaxChars.mockReturnValue(32_000);
  mockedTruncateOversizedToolResultsInSession.mockReset();
  mockedTruncateOversizedToolResultsInSession.mockResolvedValue({
    truncated: false,
@@ -420,6 +423,7 @@ export async function loadRunOverflowCompactionHarness(): Promise<{
  }));

  vi.doMock("./tool-result-truncation.js", () => ({
+    resolveLiveToolResultMaxChars: mockedResolveLiveToolResultMaxChars,
    sessionLikelyHasOversizedToolResults: mockedSessionLikelyHasOversizedToolResults,
    truncateOversizedToolResultsInSession: mockedTruncateOversizedToolResultsInSession,
  }));
--- a/src/agents/pi-embedded-runner/tool-result-context-guard.test.ts
+++ b/src/agents/pi-embedded-runner/tool-result-context-guard.test.ts
@@ -505,14 +505,15 @@ describe("installContextEngineLoopHook", () => {
    const engine = makeMockEngine({ omitAfterTurn: true, omitIngestBatch: true });
    installHook(agent, engine, 1);

-    const messages = [makeUser("first"), makeToolResult("call_1", "r1")];
+    const toolResult = makeToolResult("call_1", "r1");
+    const messages = [makeUser("first"), toolResult];
    await callTransform(agent, messages);

    expect(engine.ingest).toHaveBeenCalledTimes(1);
    expect(engine.ingest.mock.calls[0]?.[0]).toMatchObject({
      sessionId,
      sessionKey,
-      message: makeToolResult("call_1", "r1"),
+      message: toolResult,
    });
    expect(engine.assemble).toHaveBeenCalledTimes(1);
  });
--- a/src/agents/pi-tools.create-openclaw-coding-tools.adds-claude-style-aliases-schemas-without-dropping-g.test.ts
+++ b/src/agents/pi-tools.create-openclaw-coding-tools.adds-claude-style-aliases-schemas-without-dropping-g.test.ts
@@ -86,7 +86,7 @@ describe("createOpenClawCodingTools read behavior", () => {
      const result = await readTool.execute("read-cap-1", { path: "huge.txt" });
      const text = extractToolText(result);
      expect(text).toContain("line-0001");
-      expect(text).toContain("[Read output capped at 50KB for this call. Use offset=");
+      expect(text).toContain("[Read output capped at 32KB for this call. Use offset=");
      expect(text).not.toContain("line-8000");
    } finally {
      await fs.rm(tmpDir, { recursive: true, force: true });
--- a/src/auto-reply/reply/commands-context-report.test.ts
+++ b/src/auto-reply/reply/commands-context-report.test.ts
@@ -92,8 +92,8 @@ describe("buildContextReply", () => {
        omitBootstrapLimits: true,
      }),
    );
-    expect(result.text).toContain("Bootstrap max/file: 20,000 chars");
-    expect(result.text).toContain("Bootstrap max/total: 150,000 chars");
+    expect(result.text).toContain("Bootstrap max/file: 12,000 chars");
+    expect(result.text).toContain("Bootstrap max/total: 60,000 chars");
    expect(result.text).not.toContain("Bootstrap max/file: ? chars");
  });

--- a/src/gateway/server-methods/chat.directive-tags.test.ts
+++ b/src/gateway/server-methods/chat.directive-tags.test.ts
@@ -486,6 +486,13 @@ describe("chat directive tag stripping for non-streaming final payloads", () =>
    const transcriptDir = path.dirname(mockState.transcriptPath);
    const audioPath = path.join(transcriptDir, "reply.mp3");
    fs.writeFileSync(audioPath, Buffer.from([0xff, 0xfb, 0x90, 0x00]));
+    mockState.config = {
+      agents: {
+        defaults: {
+          workspace: transcriptDir,
+        },
+      },
+    };
    mockState.triggerAgentRunStart = true;
    mockState.dispatchedReplies = [
      {
@@ -506,32 +513,34 @@ describe("chat directive tag stripping for non-streaming final payloads", () =>
      expectBroadcast: false,
    });

-    const assistantUpdate = mockState.emittedTranscriptUpdates.find(
-      (update) =>
-        typeof update.message === "object" &&
-        update.message !== null &&
-        (update.message as { role?: unknown }).role === "assistant" &&
-        Array.isArray((update.message as { content?: unknown }).content) &&
-        ((update.message as { content: Array<{ type?: string }> }).content.some(
-          (block) => block?.type === "audio",
-        ) ??
-          false),
-    );
-    expect(assistantUpdate).toMatchObject({
-      message: {
-        role: "assistant",
-        idempotencyKey: "idem-agent-audio:assistant-audio",
-        content: [
-          { type: "text", text: "Audio reply" },
-          {
-            type: "audio",
-            source: {
-              type: "base64",
-              media_type: "audio/mpeg",
+    await waitForAssertion(() => {
+      const assistantUpdate = mockState.emittedTranscriptUpdates.find(
+        (update) =>
+          typeof update.message === "object" &&
+          update.message !== null &&
+          (update.message as { role?: unknown }).role === "assistant" &&
+          Array.isArray((update.message as { content?: unknown }).content) &&
+          ((update.message as { content: Array<{ type?: string }> }).content.some(
+            (block) => block?.type === "audio",
+          ) ??
+            false),
+      );
+      expect(assistantUpdate).toMatchObject({
+        message: {
+          role: "assistant",
+          idempotencyKey: "idem-agent-audio:assistant-audio",
+          content: [
+            { type: "text", text: "Audio reply" },
+            {
+              type: "audio",
+              source: {
+                type: "base64",
+                media_type: "audio/mpeg",
+              },
            },
-          },
-        ],
-      },
+          ],
+        },
+      });
    });
  });