Compare commits

..
Author SHA1 Message Date
fffa779172 fix(rerank): classify missing auth before fallback (#2059)
Missing ZEROENTROPY_API_KEY threw AIConfigError from auth resolution, which
rerank.ts recorded as reason 'unknown' — and doctor's reranker_health had no
unknown bucket, so it reported ok while every rerank silently failed open.

- gateway.rerank wraps AIConfigError from applyResolveAuth as
  RerankError(reason: 'auth') before any HTTP call.
- checkRerankerHealth warns on >=3 'unknown' failures in the 7-day window
  (covers historical pre-fix audit rows), with a ZEROENTROPY_API_KEY setup
  hint when the error summary points at a missing key.
- Tests: RerankError(auth) classification, applyReranker fail-open + audit
  reason, doctor warn on repeated unknowns.

Takeover of #2070.

Co-authored-by: maxpetrusenkoagent <maxpetrusenkoagent@users.noreply.github.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-21 14:43:03 -07:00
9 changed files with 122 additions and 59 deletions
+18
View File
@@ -1551,6 +1551,24 @@ export async function checkRerankerHealth(engine: BrainEngine): Promise<Check> {
};
}
// Historical #2059 rows were logged as `unknown` before missing reranker
// auth was classified at the gateway. Surface repeated unknowns instead of
// reporting "ok" while every rerank fails open.
const unknownFails = failures.filter((f) => f.reason === 'unknown');
if (unknownFails.length >= 3) {
const setupHint = unknownFails.some((f) => {
const summary = String(f.error_summary ?? '');
return summary.includes('ZEROENTROPY_API_KEY') || summary.toLowerCase().includes('api key');
})
? ' Fix: verify ZEROENTROPY_API_KEY and run `gbrain models doctor`.'
: '';
return {
name: 'reranker_health',
status: 'warn',
message: `${unknownFails.length} unknown reranker failure(s) in last 7 days.${setupHint}`,
};
}
return {
name: 'reranker_health',
status: 'ok',
+1 -9
View File
@@ -1743,15 +1743,7 @@ async function extractStaleFromDB(
// `page.updated_at.toISOString()` — the JS Date is ms-truncated, so the
// µs-precision DB updated_at stayed strictly greater and the page never
// cleared on Postgres. Stamping the exact value makes them equal.
//
// Version-arm floor: a page last edited BEFORE LINK_EXTRACTOR_VERSION_TS
// would otherwise be stamped below the version watermark and stay
// permanently stale (`links_extracted_at < versionTs` re-fires every run).
// Stamp max(updated_at, versionTs) — versionTs is always a past release
// date, so a concurrent edit's now() still exceeds the stamp and D4 holds.
// Tie at ms precision picks updated_at_iso (its µs ≥ versionTs's .000000).
const stampTs = new Date(page.updated_at_iso) >= new Date(versionTs) ? page.updated_at_iso : versionTs;
processedRefs.push({ slug: page.slug, source_id: page.source_id, extractedAt: stampTs });
processedRefs.push({ slug: page.slug, source_id: page.source_id, extractedAt: page.updated_at_iso });
}
// Flush NON-swallowing (CDX-4): a throw here propagates out of the sweep so
+9 -1
View File
@@ -3656,7 +3656,15 @@ export async function rerank(input: RerankInput): Promise<RerankResult[]> {
// whose request/response shape differs from ZE/llama.cpp (e.g. Voyage with
// `top_k` / `data[]`) needs separate adapter hooks in a follow-up plan.
const url = `${compat.baseURL.replace(/\/$/, '')}${tp.path ?? '/models/rerank'}`;
const auth = applyResolveAuth(recipe, cfg, 'reranker');
let auth: { apiKey?: string; headers?: Record<string, string> };
try {
auth = applyResolveAuth(recipe, cfg, 'reranker');
} catch (err) {
if (err instanceof AIConfigError) {
throw new RerankError(err.message, 'auth');
}
throw err;
}
// applyResolveAuth returns { apiKey } for Bearer-style auth (SDK's native
// path) or { headers } for custom-header providers (Azure). v0.37.6.0:
// recipes can ALSO declare default_headers (attribution etc.) which flow
+4 -13
View File
@@ -28,7 +28,7 @@ import { ensureWellFormed } from './text-safe.ts';
* OR updated_at > links_extracted_at`. It is an ISO-8601 string (NOT a number) —
* the column is TIMESTAMPTZ and the predicate binds it as `::timestamptz`.
*/
export const LINK_EXTRACTOR_VERSION_TS = '2026-07-21T00:00:00Z';
export const LINK_EXTRACTOR_VERSION_TS = '2026-05-31T00:00:00Z';
// ─── Entity references ──────────────────────────────────────────
@@ -80,10 +80,10 @@ export type LinkResolutionType = 'qualified' | 'unqualified';
* Directory prefix whitelist. These are the top-level slug dirs the extractor
* recognizes as entity references. Upstream canonical + our extensions:
* - Gbrain canonical: people, companies, meetings, concepts, deal, civic, project, source, media, yc, projects
* - Our domain extensions: tech, finance, personal, openclaw, ops (domain-organized wikis)
* - Our domain extensions: tech, finance, personal, openclaw (domain-organized wikis)
* - Our entity prefix: entities (we kept some legacy entities/projects/ pages)
*/
const DIR_PATTERN = '(?:people|companies|meetings|concepts|deal|civic|project|projects|source|media|yc|tech|finance|personal|openclaw|entities|ops)';
const DIR_PATTERN = '(?:people|companies|meetings|concepts|deal|civic|project|projects|source|media|yc|tech|finance|personal|openclaw|entities)';
/**
* Match `[Name](path)` markdown links pointing to entity directories.
@@ -865,16 +865,7 @@ export function queryBasenameIndex(idx: Map<string, string[]>, name: string): st
if (!name || typeof name !== 'string') return [];
const trimmed = name.trim();
if (!trimmed) return [];
let hit = idx.get(trimmed) ?? idx.get(trimmed.toLowerCase()) ?? idx.get(normalizeBasename(trimmed));
// Issue #2576 bug 2: path-style refs (`runbooks/2026-05-01-x`) from dirs
// outside DIR_PATTERN reach here, but normalizeBasename strips slashes
// into a garbage key (`runbooks2026-05-01-x`) that can never hit the
// tail-keyed index. Fall back to the path tail so qualified refs resolve
// by basename like everything else.
if (!hit && trimmed.includes('/')) {
const tail = trimmed.slice(trimmed.lastIndexOf('/') + 1).trim();
if (tail) hit = idx.get(tail) ?? idx.get(tail.toLowerCase()) ?? idx.get(normalizeBasename(tail));
}
const hit = idx.get(trimmed) ?? idx.get(trimmed.toLowerCase()) ?? idx.get(normalizeBasename(trimmed));
return hit ? [...hit].sort(basenameSort) : [];
}
+22
View File
@@ -154,6 +154,28 @@ describe('gateway.rerank() — happy path', () => {
describe('gateway.rerank() — error classification', () => {
beforeEach(() => configureZE());
test('missing required reranker API key → RerankError(auth) before HTTP call', async () => {
configureGateway({
reranker_model: 'zeroentropyai:zerank-2',
env: {},
});
let called = false;
__setRerankTransportForTests(async () => {
called = true;
return mockResp({ results: [{ index: 0, relevance_score: 0.5 }] });
});
try {
await rerank({ query: 'q', documents: ['d'] });
throw new Error('should have thrown');
} catch (err) {
expect(err).toBeInstanceOf(RerankError);
expect((err as RerankError).reason).toBe('auth');
expect((err as Error).message).toContain('ZEROENTROPY_API_KEY');
expect(called).toBe(false);
}
});
test('401 → auth', async () => {
__setRerankTransportForTests(async () => new Response('Unauthorized', { status: 401 }));
try {
+33
View File
@@ -2,6 +2,11 @@ import { describe, test, expect, beforeAll, afterAll, beforeEach } from 'bun:tes
import { mkdirSync, rmSync, writeFileSync } from 'fs';
import { join } from 'path';
import { tmpdir } from 'os';
import * as fs from 'node:fs';
import * as os from 'node:os';
import * as path from 'node:path';
import { withEnv } from './helpers/with-env.ts';
import { logRerankFailure } from '../src/core/rerank-audit.ts';
describe('doctor command', () => {
test('doctor module exports runDoctor', async () => {
@@ -47,6 +52,34 @@ describe('doctor command', () => {
expect(check.issues![0].action).toContain('trigger');
});
test('reranker_health warns on repeated unknown rerank failures', async () => {
const { checkRerankerHealth } = await import('../src/commands/doctor.ts');
const tmpDir = fs.mkdtempSync(path.join(os.tmpdir(), 'gbrain-rerank-doctor-'));
try {
await withEnv({ GBRAIN_AUDIT_DIR: tmpDir }, async () => {
for (let i = 0; i < 3; i++) {
logRerankFailure({
model: 'zeroentropyai:zerank-2',
reason: 'unknown',
query_hash: `unknown${i}`,
doc_count: 30,
error_summary: 'ZeroEntropy reranker requires ZEROENTROPY_API_KEY.',
});
}
const check = await checkRerankerHealth({
async getConfig(key: string): Promise<string | null> {
return key === 'search.reranker.enabled' ? 'true' : null;
},
} as any);
expect(check.status).toBe('warn');
expect(check.message).toContain('unknown');
expect(check.message).toContain('ZEROENTROPY_API_KEY');
});
} finally {
fs.rmSync(tmpDir, { recursive: true, force: true });
}
});
test('runDoctor accepts null engine for filesystem-only mode', async () => {
const { runDoctor } = await import('../src/commands/doctor.ts');
// runDoctor should accept null engine — it runs filesystem checks only.
-18
View File
@@ -209,24 +209,6 @@ describe('gbrain extract --stale', () => {
expect(usRows[0]?.eq).toBe(true);
});
test('version-arm floor: page edited BEFORE LINK_EXTRACTOR_VERSION_TS clears after --stale (issue #2576 bug 3)', async () => {
// A page whose updated_at predates the version watermark used to be
// stamped at its updated_at (< versionTs), so the version arm re-fired
// every run — permanently stale. The sweep now floors the stamp at
// versionTs. (The #1768 test above also covers this since the v0.42.x
// VERSION_TS bump moved its date below the watermark, but this pins the
// behavior explicitly so a date "repair" there can't drop coverage.)
await engine.putPage('people/alice', personPage('Alice'));
await engine.executeRaw(`UPDATE pages SET updated_at = '2000-01-01T00:00:00Z' WHERE slug = 'people/alice'`);
expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(1);
await runExtract(engine, ['--stale']);
// Pre-floor this stayed 1 forever (stamp < versionTs → version arm re-fires).
expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(0);
await runExtract(engine, ['--stale']);
expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(0);
});
test('CDX-4 (D2): a link-flush throw aborts the sweep and leaves pages UNSTAMPED', async () => {
await engine.putPage('people/alice', personPage('Alice'));
await engine.putPage('companies/acme', companyPage('Acme', '[Alice](people/alice) founded [Acme](companies/acme).'));
-18
View File
@@ -140,15 +140,6 @@ describe('extractEntityRefs', () => {
expect(wikiRefs[0].needsResolution).toBe(true);
});
test('recognizes ops/ qualified wikilinks (issue #2576 bug 2)', () => {
// `ops` was missing from DIR_PATTERN, so [[ops/...]] fell through to
// the generic 2c pass (needsResolution) instead of being a real ref.
const refs = extractEntityRefs('Deployed via [[ops/services/pointer-agent]].');
expect(refs.length).toBe(1);
expect(refs[0].slug).toBe('ops/services/pointer-agent');
expect(refs[0].needsResolution).toBeUndefined();
});
test('skips qualified-syntax tokens (those belong to 2a)', () => {
// [[wiki:topics/ai]] looks like 2a's qualified shape — even though
// it wouldn't satisfy DIR_PATTERN, 2c must not claim it either
@@ -1078,15 +1069,6 @@ describe('makeResolver — fallback chain', () => {
]);
});
test('resolveBasenameMatches: path-style ref falls back to the tail (issue #2576 bug 2)', async () => {
// normalizeBasename strips slashes, so `runbooks/2026-05-01-pointer-agent`
// used to normalize to a garbage key that never hit the tail-keyed index.
const engine = makeFakeEngineWithSlugs(['ops/changes/2026-05-01-pointer-agent']);
const r = makeResolver(engine);
expect(await r.resolveBasenameMatches!('runbooks/2026-05-01-pointer-agent'))
.toEqual(['ops/changes/2026-05-01-pointer-agent']);
});
test('resolveBasenameMatches: case-insensitive fallback', async () => {
const engine = makeFakeEngineWithSlugs(['companies/fast-weigh']);
const r = makeResolver(engine);
+35
View File
@@ -12,9 +12,14 @@
*/
import { describe, test, expect, beforeAll, afterAll } from 'bun:test';
import * as fs from 'node:fs';
import * as os from 'node:os';
import * as path from 'node:path';
import { applyReranker, type RerankerOpts } from '../../src/core/search/rerank.ts';
import { RerankError, type RerankResult } from '../../src/core/ai/gateway.ts';
import { readRecentRerankFailures } from '../../src/core/rerank-audit.ts';
import type { SearchResult } from '../../src/core/types.ts';
import { withEnv } from '../helpers/with-env.ts';
function makeResult(slug: string, score: number, chunk: string): SearchResult {
return {
@@ -160,6 +165,36 @@ describe('applyReranker — fail-open on every RerankError reason', () => {
expect(out).toEqual(results);
});
test('missing gateway reranker API key fail-opens and audits auth', async () => {
const { configureGateway } = await import('../../src/core/ai/gateway.ts');
const tmpDir = fs.mkdtempSync(path.join(os.tmpdir(), 'gbrain-rerank-search-'));
try {
await withEnv({ GBRAIN_AUDIT_DIR: tmpDir }, async () => {
configureGateway({
reranker_model: 'zeroentropyai:zerank-2',
env: {},
});
const results = [makeResult('a', 1.0, 'doc a')];
const out = await applyReranker('q', results, {
enabled: true,
topNIn: 1,
topNOut: null,
model: 'zeroentropyai:zerank-2',
});
expect(out).toEqual(results);
const failures = readRecentRerankFailures(1);
expect(failures).toHaveLength(1);
expect(failures[0]!.reason).toBe('auth');
expect(failures[0]!.error_summary).toContain('ZEROENTROPY_API_KEY');
});
} finally {
fs.rmSync(tmpDir, { recursive: true, force: true });
configureGateway({ env: { ZEROENTROPY_API_KEY: 'test-key' } });
}
});
test('fail-open on non-RerankError throw too', async () => {
const results = [makeResult('a', 1.0, 'a')];
const opts: RerankerOpts = {