Compare commits

..
Author SHA1 Message Date
f68eb9d905 fix(takes): kill propose_takes rescan loop, keep every claim, and route cycle/brainstorm models through config (#2804 #2805)
Takeover of #2804 and #2805 (stacked), rebased onto current master.

Rescan loop + dropped claims (#2804):
- Zero-claim scans now write a status='empty' sentinel row so the
  (source_id, page_slug, content_hash, prompt_version) cache hits on
  every subsequent cycle instead of re-spending the extractor LLM call
  on the same unchanged page forever.
- The idempotency index gains md5(claim_text): multi-claim pages keep
  every claim (the 4-column unique index silently dropped claims 2..N
  via ON CONFLICT DO NOTHING). proposals_inserted now counts rows that
  actually landed (RETURNING id), not insert attempts.
- Migration renumbered 123 -> 125 (master shipped v123
  configurable_fts_language and v124 after the PR was cut; the
  duplicate v123 would never have run under the version > current
  runner). Schema parity across schema.sql, pglite-schema.ts,
  schema-embedded.ts.

Model-tier honoring (#2805), reworked for current master:
- propose_takes, grade_takes, calibration_profile resolve their model
  through resolveModel(models.<phase> > models.default > env >
  getChatModel()) — per-phase config keys are now honored while the
  gateway chat model (already tier-resolved by
  reconfigureGatewayWithEngine) stays the default, preserving master's
  #2451 semantics (provider-prefixed stored ids, nested prefixes like
  openrouter:... intact) and the #2997 opts.model ?? getChatModel()
  idiom. The PR's original bare-tail-everywhere approach was dropped
  where it conflicted with #2451; grade_takes keeps the bare tail ONLY
  for its cache key / stored judge_model_id (continuity with historical
  rows), while the judge call itself gets the full resolved string —
  fixing the live label-vs-actual mismatch (call rode chat_model,
  telemetry recorded a hardcoded claude-sonnet-4-6).
- brainstorm resolves via models.brainstorm > tier chain (was
  hardcoded anthropic:claude-sonnet-4-6).
- gbrain models lists the four new per-task keys; output caps and
  context-window entries added for newer Claude models.

Tests: test/propose-takes-rescan.test.ts (new, hermetic PGLite);
config-key coverage in propose-takes/grade-takes tests; existing #2451
and gateway-chat-model tests pass unchanged.

Co-authored-by: p3ob7o <p3ob7o@users.noreply.github.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-21 14:33:19 -07:00
27 changed files with 450 additions and 360 deletions
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
+1 -1
View File
@@ -7,7 +7,7 @@
<link rel="preconnect" href="https://fonts.googleapis.com" />
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin />
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600&family=JetBrains+Mono:wght@400;500&display=swap" rel="stylesheet" />
<script type="module" crossorigin src="/admin/assets/index-CzLjRij_.js"></script>
<script type="module" crossorigin src="/admin/assets/index-CoGEje3-.js"></script>
<link rel="stylesheet" crossorigin href="/admin/assets/index-GxkWX7v3.css">
</head>
<body>
-1
View File
@@ -39,7 +39,6 @@ export const api = {
stats: () => apiFetch('/admin/api/stats'),
health: () => apiFetch('/admin/api/health-indicators'),
agents: () => apiFetch('/admin/api/agents'),
sources: () => apiFetch('/admin/api/sources'),
requests: (page = 1, qs = '') => apiFetch(`/admin/api/requests?page=${page}${qs}`),
apiKeys: () => apiFetch('/admin/api/api-keys'),
createApiKey: (name: string) => apiFetch('/admin/api/api-keys', { method: 'POST', body: JSON.stringify({ name }) }),
+1 -87
View File
@@ -23,16 +23,9 @@ interface Agent {
total_requests: number;
requests_today: number;
token_ttl: number | null;
source_id: string | null;
federated_read: string[] | null;
status: 'active' | 'revoked';
}
interface SourceOption {
id: string;
name: string | null;
}
interface ApiKey {
id: string;
name: string;
@@ -95,7 +88,6 @@ export function AgentsPage() {
<th>Name</th>
<th>Type</th>
<th>Scopes</th>
<th>Source</th>
<th>Status</th>
<th>Requests</th>
<th>Last Used</th>
@@ -116,9 +108,6 @@ export function AgentsPage() {
<span key={s} className={`badge badge-${s}`} style={{ marginRight: 4 }}>{s}</span>
))}
</td>
<td className="mono" style={{ fontSize: 12, color: 'var(--text-secondary)' }}>
{a.auth_type === 'oauth' ? (a.source_id || 'default') : 'all'}
</td>
<td>
<span className={`badge ${a.status === 'active' ? 'badge-success' : 'badge-danger'}`}>{a.status}</span>
</td>
@@ -268,29 +257,9 @@ function RegisterModal({ onClose, onRegistered }: {
Object.fromEntries(ALLOWED_SCOPES_LIST.map(s => [s, s === 'read'])) as Record<Scope, boolean>,
);
const [ttl, setTtl] = useState('86400'); // 24h default
const [redirectUris, setRedirectUris] = useState('');
const [sources, setSources] = useState<SourceOption[]>([{ id: 'default', name: 'default' }]);
const [sourceId, setSourceId] = useState('default');
const [readSources, setReadSources] = useState<Record<string, boolean>>({ default: true });
const [loading, setLoading] = useState(false);
const [error, setError] = useState('');
useEffect(() => {
api.sources()
.then((rows: SourceOption[]) => {
const nextSources = rows.length > 0 ? rows : [{ id: 'default', name: 'default' }];
const nextSourceId = nextSources.some(s => s.id === sourceId) ? sourceId : nextSources[0].id;
setSources(nextSources);
setSourceId(nextSourceId);
setReadSources(prev => {
const next = Object.fromEntries(nextSources.map(s => [s.id, Boolean(prev[s.id])])) as Record<string, boolean>;
next[nextSourceId] = true;
return next;
});
})
.catch(() => {});
}, []);
const ttlOptions = [
{ label: '1 hour', value: '3600' },
{ label: '24 hours', value: '86400' },
@@ -308,24 +277,11 @@ function RegisterModal({ onClose, onRegistered }: {
try {
// Use the CLI registration endpoint (POST to admin API)
const selectedScopes = Object.entries(scopes).filter(([, v]) => v).map(([k]) => k).join(' ');
const federatedRead = sources.map(s => s.id).filter(id => readSources[id]);
if (federatedRead.length === 0) { setError('Select at least one read source'); setLoading(false); return; }
// #1036: redirect URIs, one per line. Mirrors the CLI convention
// (src/commands/auth.ts): presence of redirect URIs implies
// authorization_code + refresh_token grants unless the user had none.
const uris = redirectUris.split('\n').map(u => u.trim()).filter(Boolean);
const res = await fetch('/admin/api/register-client', {
method: 'POST',
credentials: 'same-origin',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: name.trim(),
scopes: selectedScopes,
tokenTtl: ttl === '0' ? 315360000 : Number(ttl),
source_id: sourceId,
federated_read: federatedRead,
...(uris.length > 0 ? { redirectUris: uris, grantTypes: ['authorization_code', 'refresh_token'] } : {}),
}),
body: JSON.stringify({ name: name.trim(), scopes: selectedScopes, tokenTtl: ttl === '0' ? 315360000 : Number(ttl) }),
});
if (!res.ok) throw new Error('Registration failed');
const data = await res.json();
@@ -356,42 +312,6 @@ function RegisterModal({ onClose, onRegistered }: {
))}
</div>
</div>
<div style={{ marginBottom: 16 }}>
<label>Write Source</label>
<select value={sourceId} onChange={e => {
const id = e.target.value;
setSourceId(id);
setReadSources(p => ({ ...p, [id]: true }));
}}
style={{ width: '100%', background: 'var(--bg-secondary)', color: 'var(--text-primary)', border: '1px solid var(--border)', borderRadius: 6, padding: '6px 10px', fontSize: 14 }}>
{sources.map(s => <option key={s.id} value={s.id}>{s.name && s.name !== s.id ? `${s.id} (${s.name})` : s.id}</option>)}
</select>
</div>
<div style={{ marginBottom: 16 }}>
<label>Read Sources</label>
<div className="checkbox-group">
{sources.map(s => (
<label key={s.id} className="checkbox-label">
<input
type="checkbox"
checked={Boolean(readSources[s.id])}
onChange={e => setReadSources(p => ({ ...p, [s.id]: e.target.checked }))}
/>
{s.name && s.name !== s.id ? `${s.id} (${s.name})` : s.id}
</label>
))}
</div>
</div>
<div style={{ marginBottom: 16 }}>
<label>Redirect URIs <span style={{ color: 'var(--text-secondary)', fontWeight: 400 }}>(optional, one per line enables authorization_code flow)</span></label>
<textarea
placeholder={'https://example.com/callback'}
value={redirectUris}
onChange={e => setRedirectUris(e.target.value)}
rows={2}
style={{ width: '100%', background: 'var(--bg-secondary)', color: 'var(--text-primary)', border: '1px solid var(--border)', borderRadius: 6, padding: '6px 10px', fontSize: 13, fontFamily: 'inherit', resize: 'vertical' }}
/>
</div>
<div style={{ marginBottom: 20 }}>
<label>Token Lifetime</label>
<select value={ttl} onChange={e => setTtl(e.target.value)}
@@ -608,8 +528,6 @@ function AgentDrawer({ agent, onClose, onRevoked }: { agent: Agent; onClose: ()
client_name: agentName,
auth_type: agent.auth_type,
scope: agent.scope,
source_id: agent.source_id,
federated_read: agent.federated_read,
}, null, 2),
};
@@ -629,10 +547,6 @@ function AgentDrawer({ agent, onClose, onRevoked }: { agent: Agent; onClose: ()
<span>{(agent.scope || '').split(' ').filter(Boolean).map(s => (
<span key={s} className={`badge badge-${s}`} style={{ marginRight: 4 }}>{s}</span>
))}</span>
<span style={{ color: 'var(--text-secondary)' }}>Write Source</span>
<span className="mono">{agent.source_id || (isOAuth ? 'default' : 'all')}</span>
<span style={{ color: 'var(--text-secondary)' }}>Read Sources</span>
<span className="mono">{isOAuth ? (agent.federated_read && agent.federated_read.length > 0 ? agent.federated_read.join(', ') : (agent.source_id || 'default')) : 'all'}</span>
<span style={{ color: 'var(--text-secondary)' }}>Registered</span>
<span>{new Date(agent.created_at).toLocaleDateString()}</span>
<span style={{ color: 'var(--text-secondary)' }}>Token TTL</span>
+3 -3
View File
@@ -1,13 +1,13 @@
// AUTO-GENERATED — do not edit by hand.
// Run `bun run scripts/build-admin-embedded.ts` to regenerate.
// Source: admin/dist/ at 2026-07-21.
// Source: admin/dist/ at 2026-05-27.
//
// Bun resolves the file: imports to a path that works at runtime even
// inside a compiled binary (`bun build --compile`). The manifest maps
// the request path the express handler sees to (resolved-path, mime).
// @ts-ignore — type: 'file' is Bun ESM, not in lib.d.ts
import A_0_assets_index_CzLjRij__js from '../admin/dist/assets/index-CzLjRij_.js' with { type: 'file' };
import A_0_assets_index_CoGEje3__js from '../admin/dist/assets/index-CoGEje3-.js' with { type: 'file' };
// @ts-ignore — type: 'file' is Bun ESM, not in lib.d.ts
import A_1_assets_index_GxkWX7v3_css from '../admin/dist/assets/index-GxkWX7v3.css' with { type: 'file' };
// @ts-ignore — type: 'file' is Bun ESM, not in lib.d.ts
@@ -19,7 +19,7 @@ export interface AdminAsset {
}
export const ADMIN_ASSETS: Record<string, AdminAsset> = {
"/admin/assets/index-CzLjRij_.js": { path: A_0_assets_index_CzLjRij__js as unknown as string, mime: "application/javascript; charset=utf-8" },
"/admin/assets/index-CoGEje3-.js": { path: A_0_assets_index_CoGEje3__js as unknown as string, mime: "application/javascript; charset=utf-8" },
"/admin/assets/index-GxkWX7v3.css": { path: A_1_assets_index_GxkWX7v3_css as unknown as string, mime: "text/css; charset=utf-8" },
"/admin/index.html": { path: A_2_index_html as unknown as string, mime: "text/html; charset=utf-8" },
};
+4
View File
@@ -50,6 +50,10 @@ const PER_TASK_KEYS: Array<{ key: string; tier: ModelTier; description: string }
{ key: 'models.eval.contradictions_judge', tier: 'utility', description: 'Contradiction probe judge (v0.34 temporal-aware)' },
{ key: 'models.expansion', tier: 'utility', description: 'Query expansion for hybrid search' },
{ key: 'models.chat', tier: 'reasoning', description: 'Default `gateway.chat()` model' },
{ key: 'models.propose_takes', tier: 'reasoning', description: 'propose_takes claim extractor' },
{ key: 'models.grade_takes', tier: 'reasoning', description: 'grade_takes verdict judge' },
{ key: 'models.calibration_profile', tier: 'reasoning', description: 'Calibration profile generator' },
{ key: 'models.brainstorm', tier: 'reasoning', description: '`gbrain brainstorm` orchestrator' },
];
interface ModelEntry {
+1 -32
View File
@@ -39,7 +39,6 @@ import * as db from '../core/db.ts';
import { sqlQueryForEngine, executeRawJsonb } from '../core/sql-query.ts';
import { MinionQueue } from '../core/minions/queue.ts';
import { isRetryableError } from '../core/retry-matcher.ts';
import { normalizeAdminClientSourceScope } from '../core/admin-source-scope.ts';
import {
computeContentHash,
validateIngestionEvent,
@@ -1087,7 +1086,6 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
const oauthClients = await sql`
SELECT c.client_id as id, c.client_name as name, 'oauth' as auth_type,
c.grant_types, c.scope, c.created_at, c.token_ttl,
c.source_id, c.federated_read,
CASE WHEN c.deleted_at IS NOT NULL THEN 'revoked' ELSE 'active' END as status,
(SELECT max(created_at) FROM mcp_request_log WHERE token_name = c.client_id) as last_used_at,
(SELECT count(*)::int FROM mcp_request_log WHERE token_name = c.client_id) as total_requests,
@@ -1097,7 +1095,6 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
const legacyKeys = await sql`
SELECT a.id, a.name, 'api_key' as auth_type,
'{"bearer"}' as grant_types, 'read write admin' as scope, a.created_at, null as token_ttl,
null as source_id, null as federated_read,
CASE WHEN a.revoked_at IS NOT NULL THEN 'revoked' ELSE 'active' END as status,
a.last_used_at,
(SELECT count(*)::int FROM mcp_request_log WHERE token_name = a.name) as total_requests,
@@ -1110,17 +1107,6 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
}
});
app.get('/admin/api/sources', requireAdmin, async (_req: Request, res: Response) => {
try {
const sources = await engine.listAllSources({ includeArchived: false });
const options = sources.map(s => ({ id: s.id, name: s.name }));
if (!options.some(s => s.id === 'default')) options.unshift({ id: 'default', name: 'default' });
res.json(options);
} catch (e) {
res.status(503).json({ error: 'service_unavailable' });
}
});
// v0.38 Slice 4 — per-OAuth-client agent spend viewer. Pre-computes today's
// spend (committed + pending reservations) per client so the Agents tab
// can render a "$X / $Y today" cell. Read-side endpoint only — no mutation.
@@ -1454,8 +1440,6 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
// missing, empty) and rejects the rest with a structured 400.
const { name, tokenTtl, grantTypes, redirectUris, tokenEndpointAuthMethod } = req.body;
const rawScopes = (req.body as Record<string, unknown>).scopes ?? (req.body as Record<string, unknown>).scope;
const rawSourceId = (req.body as Record<string, unknown>).source_id ?? (req.body as Record<string, unknown>).sourceId;
const rawFederatedRead = (req.body as Record<string, unknown>).federated_read ?? (req.body as Record<string, unknown>).federatedRead;
if (!name) { res.status(400).json({ error: 'Name required' }); return; }
let scopeString: string;
try {
@@ -1486,23 +1470,8 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
});
return;
}
let sourceScope: ReturnType<typeof normalizeAdminClientSourceScope>;
try {
const sources = await engine.listAllSources({ includeArchived: false });
sourceScope = normalizeAdminClientSourceScope(
rawSourceId,
rawFederatedRead,
sources.map(s => s.id),
);
} catch (e) {
res.status(400).json({
error: 'invalid_source_scope',
message: e instanceof Error ? e.message : String(e),
});
return;
}
const result = await oauthProvider.registerClientManual(
name, grants, scopeString, uris, sourceScope.sourceId, sourceScope.federatedRead, validatedAuthMethod,
name, grants, scopeString, uris, 'default', undefined, validatedAuthMethod,
);
// Set per-client TTL if specified
if (tokenTtl && Number(tokenTtl) > 0) {
-51
View File
@@ -1,51 +0,0 @@
import { isValidSourceId } from './source-id.ts';
export interface AdminClientSourceScope {
sourceId: string;
federatedRead: string[];
}
function parseFederatedReadInput(raw: unknown, sourceId: string): string[] {
if (raw === undefined || raw === null) return [sourceId];
if (typeof raw === 'string') {
const values = raw.split(',').map(s => s.trim()).filter(Boolean);
return values.length > 0 ? values : [sourceId];
}
if (Array.isArray(raw)) {
if (!raw.every(v => typeof v === 'string')) {
throw new Error('federated_read must be an array of source ids');
}
const values = raw.map(v => v.trim()).filter(Boolean);
return values.length > 0 ? values : [sourceId];
}
throw new Error('federated_read must be a string or array');
}
export function normalizeAdminClientSourceScope(
rawSourceId: unknown,
rawFederatedRead: unknown,
availableSourceIds: string[],
): AdminClientSourceScope {
const knownSources = new Set(availableSourceIds);
const fallbackSourceId = knownSources.has('default') ? 'default' : (availableSourceIds[0] ?? 'default');
const sourceId = rawSourceId === undefined || rawSourceId === null || rawSourceId === ''
? fallbackSourceId
: rawSourceId;
if (!isValidSourceId(sourceId)) {
throw new Error(`Invalid source_id: ${JSON.stringify(sourceId)}`);
}
if (knownSources.size > 0 && !knownSources.has(sourceId)) {
throw new Error(`Unknown source_id: ${sourceId}`);
}
const federatedRead = Array.from(new Set(parseFederatedReadInput(rawFederatedRead, sourceId)));
for (const id of federatedRead) {
if (!isValidSourceId(id)) {
throw new Error(`Invalid federated_read source_id: ${JSON.stringify(id)}`);
}
if (knownSources.size > 0 && !knownSources.has(id)) {
throw new Error(`Unknown federated_read source_id: ${id}`);
}
}
return { sourceId, federatedRead };
}
+3
View File
@@ -61,7 +61,10 @@ export const MAX_OUTPUT_TOKENS_CEIL = 32_000;
* (with a readable error) instead of the provider's opaque HTTP 400.
*/
export const ANTHROPIC_OUTPUT_CAPS: Record<string, number> = {
'claude-fable-5': 64_000,
'claude-opus-4-8': 32_000,
'claude-opus-4-7': 32_000,
'claude-sonnet-5': 64_000,
'claude-sonnet-4-6': 64_000,
'claude-haiku-4-5': 64_000,
'claude-haiku-4-5-20251001': 64_000,
+9 -1
View File
@@ -32,6 +32,7 @@
*/
import type { BrainEngine } from '../engine.ts';
import { resolveModel } from '../model-config.ts';
import { chat as defaultChat, embedQuery, type ChatResult, type ChatOpts } from '../ai/gateway.ts';
import { hybridSearch, hybridSearchCached } from '../search/hybrid.ts';
import { fetchFar, type CloseRef, type FarPage } from './domain-bank.ts';
@@ -538,7 +539,14 @@ async function _runBrainstormInner(
const embedFn = opts.embedQueryFn ?? embedQuery;
// ---- Phase 0: cost preview + TTY grace ----
const modelStr = opts.modelOverride ?? 'anthropic:claude-sonnet-4-6';
// Tier-resolved (mirrors the cycle phases): honors models.brainstorm >
// models.default > models.tier.reasoning; the fallback keeps stock
// behavior identical (reasoning tier default IS claude-sonnet-4-6).
const modelStr = opts.modelOverride ?? await resolveModel(engine, {
configKey: 'models.brainstorm',
tier: 'reasoning',
fallback: 'anthropic:claude-sonnet-4-6',
});
const { aborted, estimate } = await previewCostAndWait({
profile,
model: modelStr,
+13 -3
View File
@@ -26,8 +26,8 @@
*/
import { BaseCyclePhase, type ScopedReadOpts, type BasePhaseOpts } from './base-phase.ts';
import { chat as gatewayChat } from '../ai/gateway.ts';
import { TIER_DEFAULTS } from '../model-config.ts';
import { chat as gatewayChat, getChatModel } from '../ai/gateway.ts';
import { resolveModel } from '../model-config.ts';
import { gateVoice, type VoiceGateGenerator, type VoiceGateJudge } from '../calibration/voice-gate.ts';
import { patternStatementTemplate, type PatternStatementSlots } from '../calibration/templates.ts';
// v0.41 T10 — domain widening. The aggregator module resolves the active
@@ -229,7 +229,16 @@ class CalibrationProfilePhase extends BaseCyclePhase {
): Promise<{ summary: string; details: Record<string, unknown>; status?: PhaseStatus }> {
const holder = opts.holder ?? 'garry';
const promptVersion = opts.promptVersion ?? CALIBRATION_PROFILE_PROMPT_VERSION;
const modelId = opts.model ?? TIER_DEFAULTS.reasoning;
// Resolved once (see propose-takes.ts for the chain): models.calibration_profile
// > models.default > env > the gateway's chat model (itself resolved
// through models.chat + the reasoning tier). Provider-prefixed per #2451
// — a bare id would make gateway.chat() throw "missing a provider
// prefix". Drives the generator's chat call, the budget label, and the
// persisted model_id, so the three can never disagree.
const modelId = opts.model ?? await resolveModel(engine, {
configKey: 'models.calibration_profile',
fallback: getChatModel(),
});
const gradeCompletion = opts.gradeCompletion ?? 1.0;
const patternsGenerator = opts.patternsGenerator ?? defaultPatternsGenerator;
const biasTagsGenerator = opts.biasTagsGenerator ?? defaultBiasTagsGenerator;
@@ -265,6 +274,7 @@ class CalibrationProfilePhase extends BaseCyclePhase {
scorecard,
holder,
attempt,
modelHint: modelId,
...(feedback !== undefined ? { feedback } : {}),
});
return lines.join('\n');
+22 -3
View File
@@ -36,7 +36,9 @@
import { createHash } from 'node:crypto';
import { BaseCyclePhase, type ScopedReadOpts, type BasePhaseOpts } from './base-phase.ts';
import { chat as gatewayChat } from '../ai/gateway.ts';
import { chat as gatewayChat, getChatModel } from '../ai/gateway.ts';
import { resolveModel } from '../model-config.ts';
import { splitProviderModelId } from '../model-id.ts';
import { GBrainError } from '../types.ts';
import type { OperationContext } from '../operations.ts';
import type { BrainEngine, Take, TakeResolution } from '../engine.ts';
@@ -395,7 +397,24 @@ class GradeTakesPhase extends BaseCyclePhase {
const autoResolve = opts.autoResolve ?? false; // D17 default OFF
const autoResolveThreshold = opts.autoResolveThreshold ?? 0.95; // D12 conservative
const resolvedByLabel = opts.resolvedByLabel ?? 'gbrain:grade_takes';
const judgeModelId = opts.model ?? 'claude-sonnet-4-6';
// Resolve the judge model ONCE (see propose-takes.ts for the chain —
// same label-vs-actual split fixed here: the judge call rode the
// gateway's chat_model while the grade cache key, evidence signature,
// and budget label recorded a hardcoded 4.6).
// NOTE: changing the resolved judge model invalidates the grade cache
// (judge_model_id is part of its key) — a one-time, budget-capped
// re-grade wave that is CORRECT, since the actual judge did change.
const judgeModelFull = opts.model ?? await resolveModel(engine, {
configKey: 'models.grade_takes',
fallback: getChatModel(),
});
// Bare tail for cache keys / evidence signatures / stored ids — the
// grade cache has always been keyed on bare ids; the gateway default
// resolves provider-prefixed, and normalizing preserves cache continuity
// on stock installs (no spurious re-judge wave from a prefix change). A
// genuinely different configured judge still invalidates, which is
// correct. The FULL string drives the actual judge call.
const judgeModelId = splitProviderModelId(judgeModelFull).model || judgeModelFull;
const useEnsemble = opts.useEnsemble ?? false;
const ensembleThreshold = opts.ensembleThreshold ?? 0.85;
@@ -468,7 +487,7 @@ class GradeTakesPhase extends BaseCyclePhase {
// Call the single-model judge. Errors on a single take log warning + continue.
let verdict: JudgeVerdict;
try {
verdict = await judge({ take, evidence, modelHint: opts.model });
verdict = await judge({ take, evidence, modelHint: judgeModelFull });
} catch (err) {
const msg = err instanceof Error ? err.message : String(err);
result.warnings.push(`judge failed on take ${take.id}: ${msg}`);
+58 -16
View File
@@ -5,11 +5,16 @@
* a tuned LLM extractor, writes the extracted gradeable claims to the
* `take_proposals` queue. User accepts/rejects via `gbrain takes propose`.
*
* Idempotency contract (D17 schema spec):
* The unique index on (source_id, page_slug, content_hash, prompt_version)
* means an unchanged page never re-spends LLM tokens. Bumping
* PROPOSE_TAKES_PROMPT_VERSION cleanly invalidates the cache so a tuned
* prompt re-runs proposals on every page.
* Idempotency contract (D17 schema spec; per-claim rows since migration v125):
* Every scan of a (source_id, page_slug, content_hash, prompt_version)
* tuple leaves at least one row one per extracted claim, or a single
* status='empty' sentinel when extraction yields nothing so an unchanged
* page never re-spends LLM tokens. Pre-v125 only proposal rows were
* written: a zero-claim page never entered the cache and was re-extracted
* on EVERY cycle (observed live: ~60 such pages × every cycle 1,400
* wasted extractor calls / ~$15 per day ~90% of total autopilot spend).
* Bumping PROPOSE_TAKES_PROMPT_VERSION cleanly invalidates the cache so a
* tuned prompt re-runs proposals on every page.
*
* F2 fence dedup:
* The phase reads the page's existing `<!-- gbrain:takes:begin -->` fence
@@ -40,6 +45,7 @@
import { randomUUID, createHash } from 'node:crypto';
import { BaseCyclePhase, type ScopedReadOpts, type BasePhaseOpts } from './base-phase.ts';
import { chat as gatewayChat, getChatModel } from '../ai/gateway.ts';
import { resolveModel } from '../model-config.ts';
import { writeReceipt } from '../extract/receipt-writer.ts';
import { upsertExtractRollup } from '../extract/rollup-writer.ts';
import { GBrainError } from '../types.ts';
@@ -307,6 +313,18 @@ class ProposeTakesPhase extends BaseCyclePhase {
const promptVersion = opts.promptVersion ?? PROPOSE_TAKES_PROMPT_VERSION;
const pageLimit = opts.pageLimit ?? 100;
const skipPagesWithFence = opts.skipPagesWithFence ?? false;
// Resolve the extractor model ONCE: models.propose_takes >
// models.default > GBRAIN_MODEL env > the gateway's chat model (which
// reconfigureGatewayWithEngine already resolved through models.chat +
// the reasoning tier). One resolved provider-prefixed string drives the
// actual chat call, the budget estimate, AND the stored model_id — so
// the recorded model can never disagree with the model that ran (#2451
// convention: stored ids are provider-prefixed, nested prefixes like
// openrouter:anthropic/... stay intact).
const extractorModelId = opts.model ?? await resolveModel(engine, {
configKey: 'models.propose_takes',
fallback: getChatModel(),
});
const proposalRunId = `propose-${new Date().toISOString().slice(0, 19).replace(/[-:T]/g, '')}-${randomUUID().slice(0, 8)}`;
const result: ProposeTakesResult = {
@@ -330,8 +348,6 @@ class ProposeTakesPhase extends BaseCyclePhase {
opts.reporter.start('propose_takes.pages' as never, pages.length);
}
const modelId = opts.model ?? getChatModel();
for (const page of pages) {
result.pages_scanned += 1;
this.tick(opts);
@@ -361,7 +377,7 @@ class ProposeTakesPhase extends BaseCyclePhase {
// Budget pre-check before the LLM call. Estimate: ~1500 input tokens + 500 output.
const budget = this.checkBudget({
modelId,
modelId: extractorModelId,
estimatedInputTokens: 1500,
maxOutputTokens: 500,
});
@@ -380,7 +396,7 @@ class ProposeTakesPhase extends BaseCyclePhase {
pagePath: page.slug,
pageBody: body,
existingTakes,
modelHint: opts.model,
modelHint: extractorModelId,
});
} catch (err) {
const msg = err instanceof Error ? err.message : String(err);
@@ -388,16 +404,42 @@ class ProposeTakesPhase extends BaseCyclePhase {
continue;
}
// Write proposals to take_proposals. Each row is a separate INSERT
// because the composite idempotency key is on the per-page tuple — a
// bulk UPSERT would collapse a same-page-multi-claim run into one row.
for (const p of proposals) {
// Zero-claim scans MUST still enter the idempotency cache. Pre-v125
// only proposal rows were written, so a page whose extraction yielded
// no gradeable claims never got a row for its (page, content_hash) —
// the cache check above missed on every subsequent cycle and the LLM
// call was re-spent on the same unchanged page, forever. The sentinel
// row (status='empty', empty claim_text) is invisible to the review
// queue (pending_idx is partial on status='pending'); it exists only
// so the cache check hits.
if (proposals.length === 0) {
await engine.executeRaw(
`INSERT INTO take_proposals
(source_id, page_slug, content_hash, prompt_version, proposal_run_id,
status, claim_text, kind, holder, weight, domain, dedup_against_fence_rows, model_id)
VALUES ($1, $2, $3, $4, $5, 'empty', '', 'none', 'brain', 0, NULL, NULL, $6)
ON CONFLICT (source_id, page_slug, content_hash, prompt_version, md5(claim_text)) DO NOTHING`,
[sourceId, page.slug, ch, promptVersion, proposalRunId, extractorModelId],
);
continue;
}
// Write proposals to take_proposals, one row per claim. The v125
// idempotency index includes md5(claim_text), so a same-page
// multi-claim run keeps EVERY claim — the pre-v125 four-column unique
// index made claims 2..N conflict with claim 1 and ON CONFLICT DO
// NOTHING silently dropped them (the review queue only ever saw the
// first claim of each page version). RETURNING id keeps
// proposals_inserted honest: it counts rows that actually landed,
// not insert attempts.
for (const p of proposals) {
const landed = await engine.executeRaw<{ id: number }>(
`INSERT INTO take_proposals
(source_id, page_slug, content_hash, prompt_version, proposal_run_id,
claim_text, kind, holder, weight, domain, dedup_against_fence_rows, model_id)
VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12)
ON CONFLICT (source_id, page_slug, content_hash, prompt_version) DO NOTHING`,
ON CONFLICT (source_id, page_slug, content_hash, prompt_version, md5(claim_text)) DO NOTHING
RETURNING id`,
[
sourceId,
page.slug,
@@ -410,10 +452,10 @@ class ProposeTakesPhase extends BaseCyclePhase {
p.weight,
p.domain ?? null,
JSON.stringify(existingTakes),
modelId,
extractorModelId,
],
);
result.proposals_inserted += 1;
if (landed.length > 0) result.proposals_inserted += 1;
}
}
+3
View File
@@ -58,8 +58,11 @@ const SUMMARY_SLUG_RE = /^[a-z0-9][a-z0-9\-]*(\/[a-z0-9][a-z0-9\-]*)*$/;
* resolver returns for known Anthropic aliases.
*/
const MODEL_CONTEXT_TOKENS: Record<string, number> = {
'claude-fable-5': 1_000_000,
'claude-opus-4-8': 1_000_000,
'claude-opus-4-7': 1_000_000,
'claude-opus-4-6': 1_000_000,
'claude-sonnet-5': 1_000_000,
'claude-sonnet-4-6': 200_000,
'claude-sonnet-4-5': 200_000,
'claude-haiku-4-5-20251001': 200_000,
+35
View File
@@ -5671,6 +5671,41 @@ export const MIGRATIONS: Migration[] = [
`);
},
},
{
version: 125,
name: 'take_proposals_empty_scan_sentinels_and_per_claim_rows',
// v0.42.x — kill the propose_takes rescan loop + stop dropping claims.
//
// Two defects, one schema touch:
// 1. Zero-claim scans never entered the idempotency cache (only
// proposal rows were written), so pages whose extraction yielded
// nothing were re-extracted on EVERY cycle. Observed live: ~60
// such pages per run ≈ 1,400 wasted extractor calls / ~$15 per
// day — ~90% of total autopilot LLM spend. Fix: the phase now
// writes a status='empty' sentinel row per zero-claim scan; the
// status CHECK gains the 'empty' value. Sentinels are excluded
// from the partial pending index, so the review queue never sees
// them.
// 2. The 4-column unique index collapsed a same-page multi-claim run
// to its FIRST claim: rows 2..N conflicted and ON CONFLICT DO
// NOTHING silently dropped them (verified live: exactly one row
// per (page, hash) across 3 days of runs). Fix: the idempotency
// index gains md5(claim_text) — per-claim rows, while the
// 4-column prefix still serves the per-scan cache lookup.
//
// Existing data is index-safe by construction: the old index guaranteed
// at most one row per 4-tuple, so the widened index has no duplicates
// to trip on. The DROP+ADD CONSTRAINT pair is idempotent as a unit.
idempotent: true,
sql: `
ALTER TABLE take_proposals DROP CONSTRAINT IF EXISTS take_proposals_status_check;
ALTER TABLE take_proposals ADD CONSTRAINT take_proposals_status_check
CHECK (status IN ('pending','accepted','rejected','superseded','empty'));
DROP INDEX IF EXISTS take_proposals_idempotency_idx;
CREATE UNIQUE INDEX IF NOT EXISTS take_proposals_idempotency_idx
ON take_proposals (source_id, page_slug, content_hash, prompt_version, md5(claim_text));
`,
},
];
export const LATEST_VERSION = MIGRATIONS.length > 0
+2 -2
View File
@@ -762,7 +762,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
proposed_at TIMESTAMPTZ NOT NULL DEFAULT now(),
proposal_run_id TEXT NOT NULL,
status TEXT NOT NULL DEFAULT 'pending'
CHECK (status IN ('pending','accepted','rejected','superseded')),
CHECK (status IN ('pending','accepted','rejected','superseded','empty')),
claim_text TEXT NOT NULL,
kind TEXT NOT NULL,
holder TEXT NOT NULL,
@@ -777,7 +777,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
predicted_brier_bucket_n INTEGER
);
CREATE UNIQUE INDEX IF NOT EXISTS take_proposals_idempotency_idx
ON take_proposals (source_id, page_slug, content_hash, prompt_version);
ON take_proposals (source_id, page_slug, content_hash, prompt_version, md5(claim_text));
CREATE INDEX IF NOT EXISTS take_proposals_pending_idx
ON take_proposals (source_id, status, proposed_at DESC)
WHERE status = 'pending';
+10 -4
View File
@@ -1274,8 +1274,14 @@ CREATE INDEX IF NOT EXISTS calibration_profiles_published_idx
WHERE published = true;
-- take_proposals: propose_takes phase queue. Idempotency cache via the
-- composite unique index (source_id, page_slug, content_hash, prompt_version)
-- mirrors v0.23 dream_verdicts. proposal_run_id supports --rollback by run.
-- composite unique index (source_id, page_slug, content_hash, prompt_version,
-- md5(claim_text)) the 4-column prefix is the per-scan cache key (mirrors
-- v0.23 dream_verdicts); md5(claim_text) makes rows per-claim so multi-claim
-- pages keep every claim (v125). status='empty' rows are zero-claim scan
-- sentinels: they hold the cache slot for a page version whose extraction
-- yielded nothing the phase never re-spends the LLM call on that page
-- version. Excluded from the partial pending index. proposal_run_id supports
-- --rollback by run.
CREATE TABLE IF NOT EXISTS take_proposals (
id BIGSERIAL PRIMARY KEY,
source_id TEXT NOT NULL REFERENCES sources(id) ON DELETE CASCADE,
@@ -1286,7 +1292,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
proposed_at TIMESTAMPTZ NOT NULL DEFAULT now(),
proposal_run_id TEXT NOT NULL,
status TEXT NOT NULL DEFAULT 'pending'
CHECK (status IN ('pending','accepted','rejected','superseded')),
CHECK (status IN ('pending','accepted','rejected','superseded','empty')),
claim_text TEXT NOT NULL,
kind TEXT NOT NULL,
holder TEXT NOT NULL,
@@ -1301,7 +1307,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
predicted_brier_bucket_n INTEGER
);
CREATE UNIQUE INDEX IF NOT EXISTS take_proposals_idempotency_idx
ON take_proposals (source_id, page_slug, content_hash, prompt_version);
ON take_proposals (source_id, page_slug, content_hash, prompt_version, md5(claim_text));
CREATE INDEX IF NOT EXISTS take_proposals_pending_idx
ON take_proposals (source_id, status, proposed_at DESC)
WHERE status = 'pending';
+10 -4
View File
@@ -1270,8 +1270,14 @@ CREATE INDEX IF NOT EXISTS calibration_profiles_published_idx
WHERE published = true;
-- take_proposals: propose_takes phase queue. Idempotency cache via the
-- composite unique index (source_id, page_slug, content_hash, prompt_version)
-- mirrors v0.23 dream_verdicts. proposal_run_id supports --rollback by run.
-- composite unique index (source_id, page_slug, content_hash, prompt_version,
-- md5(claim_text)) — the 4-column prefix is the per-scan cache key (mirrors
-- v0.23 dream_verdicts); md5(claim_text) makes rows per-claim so multi-claim
-- pages keep every claim (v125). status='empty' rows are zero-claim scan
-- sentinels: they hold the cache slot for a page version whose extraction
-- yielded nothing — the phase never re-spends the LLM call on that page
-- version. Excluded from the partial pending index. proposal_run_id supports
-- --rollback by run.
CREATE TABLE IF NOT EXISTS take_proposals (
id BIGSERIAL PRIMARY KEY,
source_id TEXT NOT NULL REFERENCES sources(id) ON DELETE CASCADE,
@@ -1282,7 +1288,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
proposed_at TIMESTAMPTZ NOT NULL DEFAULT now(),
proposal_run_id TEXT NOT NULL,
status TEXT NOT NULL DEFAULT 'pending'
CHECK (status IN ('pending','accepted','rejected','superseded')),
CHECK (status IN ('pending','accepted','rejected','superseded','empty')),
claim_text TEXT NOT NULL,
kind TEXT NOT NULL,
holder TEXT NOT NULL,
@@ -1297,7 +1303,7 @@ CREATE TABLE IF NOT EXISTS take_proposals (
predicted_brier_bucket_n INTEGER
);
CREATE UNIQUE INDEX IF NOT EXISTS take_proposals_idempotency_idx
ON take_proposals (source_id, page_slug, content_hash, prompt_version);
ON take_proposals (source_id, page_slug, content_hash, prompt_version, md5(claim_text));
CREATE INDEX IF NOT EXISTS take_proposals_pending_idx
ON take_proposals (source_id, status, proposed_at DESC)
WHERE status = 'pending';
-44
View File
@@ -1,44 +0,0 @@
import { describe, expect, test } from 'bun:test';
import { normalizeAdminClientSourceScope } from '../src/core/admin-source-scope.ts';
describe('admin register-client source scope normalization', () => {
const sources = ['default', 'team-a', 'team-b'];
test('defaults write and read scope to default source', () => {
expect(normalizeAdminClientSourceScope(undefined, undefined, sources)).toEqual({
sourceId: 'default',
federatedRead: ['default'],
});
});
test('accepts snake-case body fields with multiple read sources', () => {
expect(normalizeAdminClientSourceScope('team-a', ['team-a', 'team-b'], sources)).toEqual({
sourceId: 'team-a',
federatedRead: ['team-a', 'team-b'],
});
});
test('accepts comma-separated federated_read for API callers', () => {
expect(normalizeAdminClientSourceScope('team-a', 'team-b, default,team-b', sources)).toEqual({
sourceId: 'team-a',
federatedRead: ['team-b', 'default'],
});
});
test('falls back to selected write source when federated_read is empty', () => {
expect(normalizeAdminClientSourceScope('team-b', [], sources)).toEqual({
sourceId: 'team-b',
federatedRead: ['team-b'],
});
});
test('rejects malformed source ids before registration', () => {
expect(() => normalizeAdminClientSourceScope('../secret', undefined, sources)).toThrow(/Invalid source_id/);
expect(() => normalizeAdminClientSourceScope('team-a', ['snake_case'], sources)).toThrow(/Invalid federated_read/);
});
test('rejects valid-looking but unregistered source ids', () => {
expect(() => normalizeAdminClientSourceScope('ghost', undefined, sources)).toThrow(/Unknown source_id/);
expect(() => normalizeAdminClientSourceScope('team-a', ['ghost'], sources)).toThrow(/Unknown federated_read/);
});
});
+1
View File
@@ -38,6 +38,7 @@ function buildMockEngine(opts: { scorecard: TakesScorecard }): {
} {
const captured: CapturedSql[] = [];
const engine = {
async getConfig() { return null; },
kind: 'pglite',
async getScorecard() {
return opts.scorecard;
-26
View File
@@ -78,8 +78,6 @@ describe('v0.36.1.x #1077 — admin register-client supports PKCE public clients
// (under either name) from req.body. Pin the fallback pattern so the
// PKCE-fix regression contract stays load-bearing.
expect(src).toMatch(/req\.body[^;]*scopes\s*\?\?\s*[^;]*scope\b/);
expect(src).toMatch(/req\.body[^;]*source_id\s*\?\?\s*[^;]*sourceId\b/);
expect(src).toMatch(/req\.body[^;]*federated_read\s*\?\?\s*[^;]*federatedRead\b/);
// v0.41.3 (T4 atomicity fix, codex F4): admin endpoint now validates
// tokenEndpointAuthMethod via the shared validator and passes it to
// registerClientManual as a positional arg. Pre-v0.41.3 the route did
@@ -89,7 +87,6 @@ describe('v0.36.1.x #1077 — admin register-client supports PKCE public clients
// UPDATE block (the regex deliberately asserts the post-insert UPDATE
// is GONE).
expect(src).toMatch(/validateTokenEndpointAuthMethod\(tokenEndpointAuthMethod\)/);
expect(src).toMatch(/normalizeAdminClientSourceScope\(/);
expect(src).toMatch(/registerClientManual\([^)]*validatedAuthMethod[^)]*\)/);
// Regression guard: post-insert UPDATE flipping client_secret_hash to
// NULL based on a runtime check is exactly the non-atomic pattern T4
@@ -270,26 +267,3 @@ describe('v0.42.43.0 #2095 — volunteer-events sink + cycle purge wiring (struc
expect(src).toMatch(/purged_volunteer_events_count/);
});
});
describe('#1490 + #1036 — admin Register Agent form exposes source scope + redirect URIs', () => {
test('Agents.tsx RegisterModal sends source_id + federated_read in the POST body (#1490)', () => {
const src = readFileSync('admin/src/pages/Agents.tsx', 'utf8');
expect(src).toMatch(/source_id:\s*sourceId/);
expect(src).toMatch(/federated_read:\s*federatedRead/);
// The source list comes from the requireAdmin-gated endpoint, not a hardcoded 'default'.
expect(src).toMatch(/api\.sources\(\)/);
});
test('serve-http.ts serves /admin/api/sources behind requireAdmin (#1490)', () => {
const src = readFileSync('src/commands/serve-http.ts', 'utf8');
expect(src).toMatch(/app\.get\('\/admin\/api\/sources',\s*requireAdmin/);
});
test('Agents.tsx RegisterModal sends redirectUris + authorization_code grants when URIs are entered (#1036)', () => {
const src = readFileSync('admin/src/pages/Agents.tsx', 'utf8');
// One-per-line textarea split into an array...
expect(src).toMatch(/redirectUris\.split\('\\n'\)/);
// ...included in the body with the CLI's grant-type inference convention.
expect(src).toMatch(/redirectUris:\s*uris,\s*grantTypes:\s*\['authorization_code',\s*'refresh_token'\]/);
});
});
+1
View File
@@ -47,6 +47,7 @@ function buildMockEngine(opts: { takes: Take[] }): {
const captured: CapturedSql[] = [];
const resolves: CapturedResolve[] = [];
const engine = {
async getConfig() { return null; },
kind: 'pglite',
async listTakes() {
return opts.takes;
+24
View File
@@ -46,12 +46,14 @@ interface CapturedResolve {
function buildMockEngine(opts: {
takes: Take[];
cachedGrades?: Set<string>; // composite-key strings already in take_grade_cache
config?: Record<string, string>; // engine.getConfig plane (models.grade_takes etc.)
}): { engine: BrainEngine; captured: CapturedSql[]; resolves: CapturedResolve[] } {
const captured: CapturedSql[] = [];
const resolves: CapturedResolve[] = [];
const cached = opts.cachedGrades ?? new Set<string>();
const engine = {
async getConfig(key: string) { return opts.config?.[key] ?? null; },
kind: 'pglite',
async listTakes() {
return opts.takes;
@@ -224,6 +226,28 @@ describe('runPhaseGradeTakes — phase integration', () => {
expect(resolves).toHaveLength(0); // no canonical mutation
});
test('models.grade_takes config drives the judge call; cache key stays bare-tailed', async () => {
// Pre-fix the judge call rode the gateway's chat_model while the cache
// key / budget label recorded a hardcoded 'claude-sonnet-4-6'. The phase
// now resolves models.grade_takes; the judge gets the FULL string and
// the cache row keys on the bare tail (continuity with historical rows).
const takes = [buildTake({ id: 1, sinceDate: '2023-01-01' })];
const { engine, captured } = buildMockEngine({
takes,
config: { 'models.grade_takes': 'anthropic:claude-sonnet-5' },
});
const hints: Array<string | undefined> = [];
const judge: JudgeFn = async ({ modelHint }) => {
hints.push(modelHint);
return { verdict: 'correct', confidence: 0.9, reasoning: 'held' };
};
const result = await runPhaseGradeTakes(buildCtx(engine), { judge });
expect(result.status).toBe('ok');
expect(hints).toEqual(['anthropic:claude-sonnet-5']); // actual call gets the FULL string
const inserts = captured.filter(c => c.sql.includes('INSERT INTO take_grade_cache'));
expect(inserts[0]!.params[2]).toBe('claude-sonnet-5'); // judge_model_id is the bare tail
});
test('D17: auto-resolve OFF by default — even high-confidence verdict does NOT mutate takes', async () => {
const takes = [buildTake({ id: 1, sinceDate: '2023-01-01' })];
const { engine, resolves } = buildMockEngine({ takes });
-24
View File
@@ -125,30 +125,6 @@ describe('client registration', () => {
expect(client!.client_name).toBe('test-agent');
});
test('registerClientManual persists source_id and federated_read into AuthInfo', async () => {
await sql`
INSERT INTO sources (id, name) VALUES (${'team-a'}, ${'Team A'}), (${'team-b'}, ${'Team B'})
ON CONFLICT DO NOTHING
`;
const { clientId, clientSecret } = await provider.registerClientManual(
'source-scoped-test',
['client_credentials'],
'read write',
[],
'team-a',
['team-a', 'team-b'],
);
const tokens = await provider.exchangeClientCredentials(clientId, clientSecret!, 'read');
const authInfo = await provider.verifyAccessToken(tokens.access_token) as {
sourceId?: string;
allowedSources?: string[];
};
expect(authInfo.sourceId).toBe('team-a');
expect(authInfo.allowedSources).toEqual(['team-a', 'team-b']);
});
test('getClient returns undefined for unknown client', async () => {
const client = await provider.clientsStore.getClient('nonexistent');
expect(client).toBeUndefined();
+160
View File
@@ -0,0 +1,160 @@
/**
* propose_takes rescan-loop + dropped-claims regression tests (migration v125).
*
* Two live-observed defects, both fixed by the v125 schema + phase change:
*
* 1. RESCAN LOOP a page whose extraction yielded zero claims never
* entered the idempotency cache (only proposal rows were written), so
* every cycle re-spent the extractor call on the same unchanged page.
* Live impact: ~60 such pages × every cycle 1,400 wasted LLM calls
* (~$15) per day ~90% of total autopilot spend. Fix: status='empty'
* sentinel row per zero-claim scan.
*
* 2. DROPPED CLAIMS the 4-column unique index collapsed a same-page
* multi-claim run to its first claim (rows 2..N conflicted, ON CONFLICT
* DO NOTHING dropped them silently; verified live: exactly 1 row per
* (page, hash) over 3 days). Fix: idempotency index gains
* md5(claim_text).
*
* Hermetic: PGLite engine + injected extractor; no gateway, no LLM.
*/
import { describe, test, expect, beforeAll, afterAll } from 'bun:test';
import { PGLiteEngine } from '../src/core/pglite-engine.ts';
import { runPhaseProposeTakes, type ProposeTakesExtractor, type ProposedTake } from '../src/core/cycle/propose-takes.ts';
import type { OperationContext } from '../src/core/operations.ts';
let engine: PGLiteEngine;
function ctx(): OperationContext {
return {
engine,
remote: false,
config: {} as OperationContext['config'],
logger: { info() {}, warn() {}, error() {}, debug() {} } as unknown as OperationContext['logger'],
} as unknown as OperationContext;
}
/** Extractor stub that counts invocations per page slug. */
function countingExtractor(
claimsBySlug: Record<string, ProposedTake[]>,
): { extractor: ProposeTakesExtractor; calls: string[] } {
const calls: string[] = [];
const extractor: ProposeTakesExtractor = async ({ pagePath }) => {
calls.push(pagePath);
return claimsBySlug[pagePath] ?? [];
};
return { extractor, calls };
}
beforeAll(async () => {
engine = new PGLiteEngine();
await engine.connect({});
await engine.initSchema();
await engine.putPage('notes/zero-claims', {
type: 'note',
title: 'pure narrative',
compiled_truth: 'A quiet walk in the park. Nothing opinionated happened at all today.',
});
await engine.putPage('notes/three-claims', {
type: 'note',
title: 'opinionated',
compiled_truth: 'I bet acme-example wins the market. widget-co will struggle. fund-a is overexposed.',
});
});
afterAll(async () => {
await engine.disconnect();
});
describe('rescan loop — zero-claim scans enter the cache', () => {
test('second run cache-hits: extractor is NOT called again on unchanged pages', async () => {
const claims = {
'notes/three-claims': [
{ claim_text: 'acme-example wins the market', kind: 'bet' as const, holder: 'brain', weight: 0.7 },
{ claim_text: 'widget-co will struggle', kind: 'take' as const, holder: 'brain', weight: 0.6 },
{ claim_text: 'fund-a is overexposed', kind: 'take' as const, holder: 'brain', weight: 0.55 },
],
};
const first = countingExtractor(claims);
const r1 = await runPhaseProposeTakes(ctx(), { extractor: first.extractor });
expect(r1.status).toBe('ok');
// Both pages extracted on the first pass.
expect(first.calls).toContain('notes/zero-claims');
expect(first.calls).toContain('notes/three-claims');
const second = countingExtractor(claims);
const r2 = await runPhaseProposeTakes(ctx(), { extractor: second.extractor });
expect(r2.status).toBe('ok');
// THE regression: pre-fix the zero-claim page missed the cache every
// run and was re-extracted here. (Run 1's receipt page legitimately
// appears once — it's a new page — and its zero-claim scan now caches
// too; pre-fix, receipts re-scanned forever as well.)
expect(second.calls).not.toContain('notes/zero-claims');
expect(second.calls).not.toContain('notes/three-claims');
// Run 2 inserted nothing, so no new receipt page exists: run 3 must be
// fully quiescent — zero extractor calls, zero cache misses.
const third = countingExtractor(claims);
const r3 = await runPhaseProposeTakes(ctx(), { extractor: third.extractor });
expect(r3.status).toBe('ok');
expect(third.calls).toEqual([]);
expect((r3.details as Record<string, unknown>).cache_misses).toBe(0);
});
test('zero-claim scan wrote an "empty" sentinel invisible to the pending queue', async () => {
const sentinel = await engine.executeRaw<{ status: string; claim_text: string }>(
`SELECT status, claim_text FROM take_proposals WHERE page_slug = 'notes/zero-claims'`,
[],
);
expect(sentinel.length).toBe(1);
expect(sentinel[0].status).toBe('empty');
expect(sentinel[0].claim_text).toBe('');
const pending = await engine.executeRaw<{ n: number }>(
`SELECT COUNT(*)::int AS n FROM take_proposals WHERE page_slug = 'notes/zero-claims' AND status = 'pending'`,
[],
);
expect(Number(pending[0].n)).toBe(0);
});
});
describe('dropped claims — per-claim rows survive the idempotency index', () => {
test('a 3-claim page stores 3 rows and reports an honest inserted count', async () => {
const rows = await engine.executeRaw<{ claim_text: string }>(
`SELECT claim_text FROM take_proposals WHERE page_slug = 'notes/three-claims' AND status = 'pending' ORDER BY id`,
[],
);
// Pre-fix the 4-column unique index kept only the FIRST claim.
expect(rows.length).toBe(3);
expect(rows.map(r => r.claim_text)).toEqual([
'acme-example wins the market',
'widget-co will struggle',
'fund-a is overexposed',
]);
});
test('content change re-extracts and stores the new version separately', async () => {
await engine.putPage('notes/zero-claims', {
type: 'note',
title: 'pure narrative',
compiled_truth: 'Updated: I now believe acme-example is undervalued and will re-rate within a year.',
});
const claims = {
'notes/zero-claims': [
{ claim_text: 'acme-example is undervalued', kind: 'take' as const, holder: 'brain', weight: 0.6 },
],
};
const run = countingExtractor(claims);
const r = await runPhaseProposeTakes(ctx(), { extractor: run.extractor });
expect(r.status).toBe('ok');
// Changed page re-extracts; the unchanged 3-claim page stays cached.
expect(run.calls).toEqual(['notes/zero-claims']);
expect((r.details as Record<string, unknown>).proposals_inserted).toBe(1);
const all = await engine.executeRaw<{ status: string }>(
`SELECT status FROM take_proposals WHERE page_slug = 'notes/zero-claims' ORDER BY id`,
[],
);
// Old hash's sentinel + new hash's pending claim coexist.
expect(all.map(r2 => r2.status).sort()).toEqual(['empty', 'pending']);
});
});
+33 -1
View File
@@ -41,12 +41,16 @@ interface CapturedSql {
function buildMockEngine(opts: {
pages: Page[];
existingProposals?: Set<string>; // composite-key strings already in take_proposals
config?: Record<string, string>; // engine.getConfig plane (models.tier.* etc.)
}): { engine: BrainEngine; captured: CapturedSql[] } {
const captured: CapturedSql[] = [];
const existing = opts.existingProposals ?? new Set<string>();
const engine = {
kind: 'pglite',
async getConfig(key: string) {
return opts.config?.[key] ?? null;
},
async listPages() {
return opts.pages;
},
@@ -59,7 +63,12 @@ function buildMockEngine(opts: {
if (existing.has(key)) return [{ id: 1 } as unknown as T];
return [];
}
// INSERT — return nothing
// INSERT ... RETURNING id — emulate a successful insert so the
// honest proposals_inserted counter (counts RETURNING rows, not
// attempts) sees the row land. Conflicted inserts would return [].
if (sql.includes('INSERT INTO take_proposals') && sql.includes('RETURNING id')) {
return [{ id: 1 } as unknown as T];
}
return [];
},
} as unknown as BrainEngine;
@@ -267,6 +276,29 @@ describe('runPhaseProposeTakes — phase integration', () => {
expect(inserts[0]!.params[9]).toBe('market'); // domain
});
test('extractor model resolves through models.propose_takes config', async () => {
// Pre-fix the phase's only model knob was the gateway chat model — there
// was no per-phase config key. The phase now resolves once via
// resolveModel(models.propose_takes > models.default > env > gateway
// chat model); the extractor hint and the stored model_id must both
// reflect the configured override (full provider-prefixed string, #2451).
const pages = [buildPage({ slug: 'wiki/concepts/tier-routing', body: 'Tier-routed models will win.' })];
const { engine, captured } = buildMockEngine({
pages,
config: { 'models.propose_takes': 'anthropic:claude-sonnet-5' },
});
const seen: Array<string | undefined> = [];
const extractor: ProposeTakesExtractor = async ({ modelHint }) => {
seen.push(modelHint);
return [{ claim_text: 'tier-routed models win', kind: 'bet', holder: 'brain', weight: 0.7 }];
};
const result = await runPhaseProposeTakes(buildCtx(engine), { extractor });
expect(result.status).toBe('ok');
expect(seen).toEqual(['anthropic:claude-sonnet-5']); // chat call gets the FULL string
const inserts = captured.filter(c => c.sql.includes('INSERT INTO take_proposals'));
expect(inserts[0]!.params[11]).toBe('anthropic:claude-sonnet-5'); // stored model_id matches the call
});
test('cache hit: page already in take_proposals is skipped', async () => {
const body = 'A page that was already processed.';
const pages = [buildPage({ slug: 'wiki/old-page', body })];