AStack/OpenClaw readiness follow-up evidence for this PR:

• Branch head: c4fde033953b466431c224c20892414595216d66
• Base checked: upstream master at 058fe695756ed16e43916d907af3845338430156 (0.26.7)
• Local PR worktree verification already passed:
  • bun run src/cli.ts doctor --fast --json | jq ' .checks[] | select(.name=="resolver_health") ' -> resolver_health status ok, 39 skills, all reachable\n - bun run src/cli.ts routing-eval --skills-dir skills --json -> ok: true, totalCases: 58, passed: 58, missed: 0, ambiguous: 0, falsePositives: 0\n - isolated tests with clean HOME/env: bun test test/resolver.test.ts test/routing-eval.test.ts test/check-resolvable.test.ts test/check-resolvable-cli.test.ts -> 161 pass, 0 fail\n\nWhy this matters for downstream readiness: current AStack/OpenClaw runtime on GBrain 0.26.6 still reports doctor warnings from 37 resolver routing misses. This PR is the upstream path that lets downstream move from PASS_WITH_CONCERNS to doctor-green without carrying an AStack-local resolver fixture patch.

Rebased onto upstream master 9c2dc4cd544cd8013e0eee7a6ffb8536d3c2f13a / 0.26.8 and force-with-lease pushed. New head: 2a52cf4fc9769a3a3303c4051a20af3e5db0baa3.

Targeted verification after rebase:

• HOME=$(mktemp -d /tmp/gbrain-test-home.XXXXXX) bun test test/check-resolvable.test.ts test/check-resolvable-cli.test.ts test/doctor.test.ts test/routing-eval.test.ts test/routing-eval-cli.test.ts --timeout 30000 -> 100 pass, 0 fail
• bun run typecheck -> pass

Note: the temporary HOME was required because the default local agent environment has ~/.openclaw/workspace, which changes resolveSkillsDir auto-detection and is unrelated to this PR.

Rebased onto upstream master ee9ceb327a39b0c705ee945c6cfe821de11d34ed / 0.27.0 and force-with-lease pushed. New head: 7026ea2e851c15e9a1489b6df50c5b0d485d2c44.\n\nTargeted verification after rebase:\n- HOME=/tmp/gbrain-test-home.PD2GqO bun run src/cli.ts doctor --fast --json -> resolver_health status ok, 39 skills, all reachable\n- HOME=/tmp/gbrain-test-home.djTORw bun run src/cli.ts routing-eval --skills-dir skills --json -> ok: true, totalCases: 58, passed: 58, missed: 0, ambiguous: 0, falsePositives: 0\n- HOME=/tmp/gbrain-test-home.7NF3qI bun test test/check-resolvable.test.ts test/check-resolvable-cli.test.ts test/doctor.test.ts test/routing-eval.test.ts test/routing-eval-cli.test.ts --timeout 30000 -> 100 pass, 0 fail\n- bun run typecheck -> pass\n\nControl check: upstream master 0.27.0 still reports resolver_health warnings and routing-eval passed: 21, missed: 37, so this PR remains needed for the downstream GBrain/OpenClaw FULL PASS doctor gate.

Live downstream evidence from AStack/OpenClaw dogfood cut on 2026-05-05:

• Consumed in custom branch ArshyaAI/gbrain@de11d4c858b1a880931c008fbee3ceef92a8329a together with PRs fix: return clean auth failures for invalid MCP bearer tokens #620 and fix: embed stale chunks by page identity #626.
• Runtime gbrain doctor --json: status=healthy, health_score=100, no warning checks.
• Candidate suite: 168 pass, 28 skip, 0 fail; bun run typecheck passed.
• Direct GBrain, Remote MCP, local Codex, local Claude Code, and OpenClaw-mediated canaries all passed against the custom cut.

This PR remains one of the upstream-clean blockers for AStack FULL PASS; current dogfood status is intentionally custom/non-upstream-clean until merged and consumed.

2026-05-06 maintainer merge packet / downstream readiness refresh:

• PR state checked today: OPEN, MERGEABLE, no checks/reviews blocking from GitHub metadata.
• AStack/OpenClaw live dogfood cut still consumes this PR via custom branch ArshyaAI/gbrain@de11d4c858b1a880931c008fbee3ceef92a8329a.
• Fresh gates: runtime gbrain doctor --json is status=healthy, health_score=100, no warning checks; Remote MCP/OAuth, direct GBrain, local Codex, local Claude Code, runtime risk logs, shell quota guard, and scheduler burn-in all pass.
• Scheduler burn-in is now complete: 29.0h/24h, no new dead jobs after cutoff 2026-05-05T05:55:53Z.
• Remaining FULL PASS blockers are now upstream-clean only: merge/consume fix: align resolver routing fixtures #619, fix: return clean auth failures for invalid MCP bearer tokens #620, fix: embed stale chunks by page identity #626, then move runtime/Docker pins from custom cut back to upstream master.

Why this PR matters: it removes the resolver fixture custom-cut dependency and makes resolver doctor health upstream-owned instead of AStack-owned.

Thanks for this contribution — and apologies for the slow triage. We did a full pass over the entire PR backlog. gbrain has moved fast, and the maintainer's larger "cathedral" rewrites have superseded a big share of community PRs: the AI gateway + recipes + user_provided_models system replaced almost all individual provider PRs; #1805 fixed the whole Postgres module-singleton class; #1542 unified the type taxonomy; #1657 the retrieval path; #1802 the doctor; and so on.

We're closing this one in that cleanup — either the fix already landed on master, it duplicates another PR or merged change, or it's outside the current merge bar. Where a closed PR carried a genuinely valuable idea, we've recorded it in docs/designs/COMMUNITY_IDEAS.md so nothing good is lost (a few may graduate into TODOs).

Please don't read the close as a judgment of the work — thank you for contributing. If you believe the underlying issue is still live on the latest master, reopen with a quick note and we'll take another look. 🙏