fix(dream): stamp incremental extraction watermark (#2636)

The Dream cycle disables sync's inline extraction and routes changed
slugs through extractForSlugs, which flushed link/timeline batches but
never stamped links_extracted_at — so incrementally extracted pages
stayed permanently visible to `extract --stale` / doctor.

Collect processedRefs per successfully processed page and stamp them
via stampExtracted (best-effort) after both batch flushes, non-dry-run
mode 'all' only. Source-id threading from the original PR #2637 already
landed on master via #1503/#1747, so this rebase carries only the
missing watermark stamp plus regression tests.

Takeover of #2637.

Co-authored-by: JavanC <JavanC@users.noreply.github.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

src/cli.ts

@@ -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 /

src/commands/extract.ts

@@ -1025,6 +1025,10 @@ async function extractForSlugs(
   let linksCreated = 0;
   let timelineCreated = 0;
   let pagesProcessed = 0;
+  // #2636: successfully processed pages get their extraction watermark
+  // stamped after the final flush (mode 'all' only — a partial-mode run
+  // hasn't done the full extraction the watermark asserts).
+  const processedRefs: Array<{ slug: string; source_id: string }> = [];
 
   // Issue #972: read the basename flag once per extract run.
   const globalBasename = await isGlobalBasenameEnabled(engine);
@@ -1113,6 +1117,7 @@ async function extractForSlugs(
         }
 
         pagesProcessed++;
+        if (!dryRun) processedRefs.push({ slug, source_id: sourceId ?? 'default' });
       } catch { /* skip unreadable */ }
       progress.tick(1);
     },
@@ -1120,6 +1125,13 @@ async function extractForSlugs(
 
   await flushLinks();
   await flushTimeline();
+  // #2636: the Dream cycle disables sync's inline extraction and routes
+  // changed slugs through this incremental path — without a stamp here,
+  // those pages never get links_extracted_at and stay permanently visible
+  // to `extract --stale` / doctor. Stamp only after BOTH batches flushed.
+  if (!dryRun && mode === 'all') {
+    await stampExtracted(engine, processedRefs);
+  }
   progress.finish();
 
   if (!jsonMode) {

src/core/source-resolver.ts

@@ -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:

test/extract-incremental.test.ts

@@ -60,6 +60,51 @@ async function seedPage(slug: string, body: string): Promise<void> {
 }
 
 describe('runExtractCore — incremental cycle path (#417)', () => {
+  test('Dream incremental all-mode stamps the source-scoped extraction watermark (#2636)', async () => {
+    await engine.executeRaw(
+      `INSERT INTO sources (id, name, local_path) VALUES ($1, $2, $3)`,
+      ['repo-a', 'repo-a', tempDir],
+    );
+    await engine.putPage('people/alice-example', {
+      type: 'person',
+      title: 'alice-example',
+      compiled_truth: '# alice',
+      timeline: '',
+      frontmatter: {},
+      content_hash: 'h',
+    }, { sourceId: 'repo-a' });
+    writeFileSync(join(tempDir, 'people/alice-example.md'), '# alice');
+
+    await runExtractCore(engine as unknown as BrainEngine, {
+      mode: 'all',
+      dir: tempDir,
+      slugs: ['people/alice-example'],
+      sourceId: 'repo-a',
+    });
+
+    const rows = await engine.executeRaw<{ links_extracted_at: string | null }>(
+      `SELECT links_extracted_at FROM pages WHERE slug = $1 AND source_id = $2`,
+      ['people/alice-example', 'repo-a'],
+    );
+    expect(rows[0]?.links_extracted_at).not.toBeNull();
+    expect(await engine.countStalePagesForExtraction({ sourceId: 'repo-a' })).toBe(0);
+  });
+
+  test('Dream incremental dry-run does NOT stamp the watermark', async () => {
+    await seedPage('people/alice-example', '# alice');
+    await runExtractCore(engine as unknown as BrainEngine, {
+      mode: 'all',
+      dir: tempDir,
+      slugs: ['people/alice-example'],
+      dryRun: true,
+    });
+    const rows = await engine.executeRaw<{ links_extracted_at: string | null }>(
+      `SELECT links_extracted_at FROM pages WHERE slug = $1`,
+      ['people/alice-example'],
+    );
+    expect(rows[0]?.links_extracted_at ?? null).toBeNull();
+  });
+
   test('1. slugs: [] returns immediately with zero counts (early-return path)', async () => {
     await seedPage('people/alice-example', '# alice');
     const result = await runExtractCore(engine as unknown as BrainEngine, {

test/thin-client-source-scope.test.ts

@@ -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();
-    });
-  });
-});