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
8 changed files with 117 additions and 234 deletions
-64
View File
@@ -24,7 +24,6 @@ import type { GBrainConfig } from './core/config.ts';
import type { AIGatewayConfig } from './core/ai/types.ts';
import type { BrainEngine } from './core/engine.ts';
import { operations, OperationError } from './core/operations.ts';
import { resolveSourceIdEngineFree } from './core/source-resolver.ts';
import { formatVolunteeredPage } from './core/context/volunteer.ts';
import type { Operation, OperationContext } from './core/operations.ts';
import { shouldForceExitAfterMain, finishCliTeardown, flushThenExit, currentExitCode, setCliExitVerdict } from './core/cli-force-exit.ts';
@@ -383,15 +382,6 @@ async function main() {
if (op.localOnly) {
refuseThinClient(command, cfgPre!.remote_mcp!.mcp_url);
}
// #2098: the local path resolves --source / GBRAIN_SOURCE / .gbrain-source
// inside makeContext (ctx.sourceId), which this route never reaches — so
// scope must be mapped onto the op's source_id wire param before the call.
try {
applyThinClientSourceScope(op, params);
} catch (e: unknown) {
console.error(e instanceof Error ? e.message : String(e));
process.exit(1);
}
await runThinClientRouted(op, params, cfgPre!, cliOpts);
return;
}
@@ -812,60 +802,6 @@ export function parseOpArgs(op: Operation, args: string[]): Record<string, unkno
return params;
}
/**
* #2098: thin-client source scoping. Locally, --source / GBRAIN_SOURCE /
* .gbrain-source resolve to ctx.sourceId in makeContext; the thin-client
* route short-circuits before that, so `gbrain query --source X` against a
* remote brain silently searched unscoped. This runs the engine-free tiers
* (flag → env → dotfile; the DB-backed tiers can't run without an engine —
* the server's grant scoping covers the rest) and maps the result onto the
* op's `source_id` wire param.
*
* Ops that declare their OWN `source` param (facts add, etc.) are left
* untouched — their --source is an op param, not scope. An explicit --source
* on an op with no source_id wire param throws (loud beats silent drop);
* ambient env/dotfile scope with nowhere to send it is ignored, matching the
* pre-fix behavior for non-scopeable ops. Exported for tests.
*/
// Ops whose `source_id` wire param is NOT read-scope semantics: get_skill's
// source_id flips the lookup from host catalog to brain-resident-pack
// (getResidentSkillDetail). Ambient env/dotfile scope must never leak into
// these; an explicit --source-id still passes through untouched above.
const NON_SCOPE_SOURCE_ID_OPS = new Set(['get_skill']);
export function applyThinClientSourceScope(
op: Operation,
params: Record<string, unknown>,
cwd?: string,
): void {
if ('source' in op.params) return; // the op owns --source; not a scope flag
const explicit = typeof params.source === 'string' && params.source.length > 0
? (params.source as string)
: null;
delete params.source; // never a wire param on these ops — don't leak it
// Explicit per-call scope already on the wire wins over ambient tiers.
if (params.source_id !== undefined || params.all_sources === true) {
if (explicit) {
throw new Error('Pass either --source or --source-id/--all-sources, not both.');
}
return;
}
const resolved = resolveSourceIdEngineFree(explicit, cwd);
if (!resolved) return;
if (!('source_id' in op.params) || NON_SCOPE_SOURCE_ID_OPS.has(op.name)) {
if (explicit) {
const hint = NON_SCOPE_SOURCE_ID_OPS.has(op.name)
? `(its source_id parameter is not a scope filter; pass --source-id explicitly if you mean it)`
: `(the remote op has no source_id parameter; the server scopes it to your grant)`;
throw new Error(
`gbrain ${op.cliHints?.name || op.name} does not accept --source on a thin-client install ${hint}.`,
);
}
return; // ambient env/dotfile scope with nowhere to send it
}
params.source_id = resolved;
}
async function makeContext(engine: BrainEngine, params: Record<string, unknown>): Promise<OperationContext> {
// v0.31.8 (D11): resolve sourceId via the canonical 6-tier chain. Honors
// --source / GBRAIN_SOURCE / .gbrain-source / path-match / brain default /
+18
View File
@@ -1551,6 +1551,24 @@ export async function checkRerankerHealth(engine: BrainEngine): Promise<Check> {
};
}
// Historical #2059 rows were logged as `unknown` before missing reranker
// auth was classified at the gateway. Surface repeated unknowns instead of
// reporting "ok" while every rerank fails open.
const unknownFails = failures.filter((f) => f.reason === 'unknown');
if (unknownFails.length >= 3) {
const setupHint = unknownFails.some((f) => {
const summary = String(f.error_summary ?? '');
return summary.includes('ZEROENTROPY_API_KEY') || summary.toLowerCase().includes('api key');
})
? ' Fix: verify ZEROENTROPY_API_KEY and run `gbrain models doctor`.'
: '';
return {
name: 'reranker_health',
status: 'warn',
message: `${unknownFails.length} unknown reranker failure(s) in last 7 days.${setupHint}`,
};
}
return {
name: 'reranker_health',
status: 'ok',
+9 -1
View File
@@ -3656,7 +3656,15 @@ export async function rerank(input: RerankInput): Promise<RerankResult[]> {
// whose request/response shape differs from ZE/llama.cpp (e.g. Voyage with
// `top_k` / `data[]`) needs separate adapter hooks in a follow-up plan.
const url = `${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
-27
View File
@@ -160,33 +160,6 @@ export async function resolveSourceId(
return 'default';
}
/**
* Engine-free tiers (1-3) of the resolution chain: explicit flag →
* GBRAIN_SOURCE env → .gbrain-source dotfile walk. Used by the thin-client
* CLI path (#2098), which has no local engine to run tiers 4-6 or
* assertSourceExists against — the remote server enforces existence + grant.
* Returns null when no engine-free tier fires.
*/
export function resolveSourceIdEngineFree(
explicit: string | null | undefined,
cwd: string = process.cwd(),
): string | null {
if (explicit) {
if (!SOURCE_ID_RE.test(explicit)) {
throw new Error(`Invalid --source value "${explicit}". Must match [a-z0-9-]{1,32}.`);
}
return explicit;
}
const env = process.env.GBRAIN_SOURCE;
if (env && env.length > 0) {
if (!SOURCE_ID_RE.test(env)) {
throw new Error(`Invalid GBRAIN_SOURCE value "${env}". Must match [a-z0-9-]{1,32}.`);
}
return env;
}
return readDotfileWalk(cwd);
}
/**
* Returns the id of the SINGLE registered non-default source with a
* local_path, when exactly one such row exists. Returns null when:
+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.
+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 = {
-142
View File
@@ -1,142 +0,0 @@
/**
* #2098: thin-client routing dropped --source / GBRAIN_SOURCE / .gbrain-source.
*
* The local CLI path resolves source scope in makeContext (ctx.sourceId); the
* thin-client route short-circuits before that and sent params verbatim, so
* `gbrain query --source X` against a remote brain silently searched unscoped
* (the server op ignores the unknown `source` key).
*
* applyThinClientSourceScope runs the engine-free tiers (flag env dotfile)
* and maps the result onto the op's `source_id` wire param. These tests fail
* without the fix (params.source_id stays undefined / params.source leaks).
*/
import { describe, test, expect } from 'bun:test';
import { mkdtempSync, rmSync, writeFileSync } from 'fs';
import { join } from 'path';
import { tmpdir } from 'os';
import { applyThinClientSourceScope, parseOpArgs } from '../src/cli.ts';
import { operationsByName } from '../src/core/operations.ts';
import { withEnv } from './helpers/with-env.ts';
const queryOp = operationsByName.query;
describe('applyThinClientSourceScope (#2098)', () => {
test('--source maps onto the query op wire param source_id', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const params = parseOpArgs(queryOp, ['find things', '--source', 'wiki']);
expect(params.source).toBe('wiki'); // pre-fix state: wrong key
applyThinClientSourceScope(queryOp, params, '/');
expect(params.source_id).toBe('wiki');
expect('source' in params).toBe(false); // never leaks the unknown key
});
});
test('GBRAIN_SOURCE env tier fires when no flag is passed', async () => {
await withEnv({ GBRAIN_SOURCE: 'gstack' }, () => {
const params = parseOpArgs(queryOp, ['find things']);
applyThinClientSourceScope(queryOp, params, '/');
expect(params.source_id).toBe('gstack');
});
});
test('.gbrain-source dotfile tier fires when flag and env are absent', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const tmp = mkdtempSync(join(tmpdir(), 'gbrain-thin-scope-'));
try {
writeFileSync(join(tmp, '.gbrain-source'), 'essays\n');
const params = parseOpArgs(queryOp, ['find things']);
applyThinClientSourceScope(queryOp, params, tmp);
expect(params.source_id).toBe('essays');
} finally {
rmSync(tmp, { recursive: true, force: true });
}
});
});
test('explicit --source-id on the wire wins over ambient env scope', async () => {
await withEnv({ GBRAIN_SOURCE: 'gstack' }, () => {
const params = parseOpArgs(queryOp, ['find things', '--source-id', 'wiki']);
applyThinClientSourceScope(queryOp, params, '/');
expect(params.source_id).toBe('wiki');
});
});
test('--source together with --source-id is rejected loudly', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const params = parseOpArgs(queryOp, ['q', '--source', 'a', '--source-id', 'b']);
expect(() => applyThinClientSourceScope(queryOp, params, '/')).toThrow(/not both/);
});
});
test('invalid --source value is rejected loudly', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const params = parseOpArgs(queryOp, ['q', '--source', 'Bad_Value!']);
expect(() => applyThinClientSourceScope(queryOp, params, '/')).toThrow(/Invalid --source/);
});
});
test('--source on an op with no source_id wire param errors instead of silently dropping', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const op = operationsByName.add_tag;
expect('source_id' in op.params).toBe(false);
const params = { slug: 'x', tag: 'y', source: 'wiki' };
expect(() => applyThinClientSourceScope(op, params, '/')).toThrow(/--source/);
});
});
test('ambient env scope on an op with no source_id wire param is ignored (no throw)', async () => {
await withEnv({ GBRAIN_SOURCE: 'wiki' }, () => {
const op = operationsByName.add_tag;
const params: Record<string, unknown> = { slug: 'x', tag: 'y' };
applyThinClientSourceScope(op, params, '/');
expect(params.source_id).toBeUndefined();
});
});
test('ops that declare their OWN source param are left untouched', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const op = operationsByName.put_raw_data;
expect('source' in op.params).toBe(true);
const params: Record<string, unknown> = { slug: 'x', source: 'crustdata', data: {} };
applyThinClientSourceScope(op, params, '/');
expect(params.source).toBe('crustdata');
expect(params.source_id).toBeUndefined();
});
});
test('get_skill: ambient scope never leaks into its non-scope source_id param', async () => {
await withEnv({ GBRAIN_SOURCE: 'wiki' }, () => {
const op = operationsByName.get_skill;
expect('source_id' in op.params).toBe(true); // has the param, but it is a mode switch
const params: Record<string, unknown> = { name: 'ingest' };
applyThinClientSourceScope(op, params, '/');
expect(params.source_id).toBeUndefined(); // would flip host catalog → brain-pack lookup
});
});
test('get_skill: explicit --source errors instead of masquerading as --source-id', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const op = operationsByName.get_skill;
const params: Record<string, unknown> = { name: 'ingest', source: 'wiki' };
expect(() => applyThinClientSourceScope(op, params, '/')).toThrow(/--source-id/);
});
});
test('get_skill: explicit --source-id passes through untouched', async () => {
await withEnv({ GBRAIN_SOURCE: 'gstack' }, () => {
const op = operationsByName.get_skill;
const params: Record<string, unknown> = { name: 'ingest', source_id: 'wiki' };
applyThinClientSourceScope(op, params, '/');
expect(params.source_id).toBe('wiki');
});
});
test('no scope from any tier leaves params unchanged', async () => {
await withEnv({ GBRAIN_SOURCE: undefined }, () => {
const params = parseOpArgs(queryOp, ['find things']);
applyThinClientSourceScope(queryOp, params, '/');
expect(params.source_id).toBeUndefined();
});
});
});