Best possible solution:

Land the narrow ClawRouter plugin fix after removing the release-owned changelog entry and adding redacted real-run proof for auth-profile model resolution and fail-closed credential behavior.

Do we have a high-confidence way to reproduce the issue?

Yes from source, not from a live run: current main can discover ClawRouter models with an auth-profile credential, but runtime model resolution has no ClawRouter dynamic hook to refresh those catalog rows for an auth-profile-only setup.

Is this the best way to solve the issue?

Yes, with one cleanup: using provider-local prepareDynamicModel/resolveDynamicModel is the narrowest maintainable seam for this bug, but the release-owned CHANGELOG edit should be dropped.

Full review comments:

• [P3] Remove the release-owned changelog entry — CHANGELOG.md:30
CHANGELOG.md is release-generated and repository policy says normal PRs should not edit it. Keeping this hand-written entry can drift from release generation; keep the release-note context in the PR body or squash message instead.
  Confidence: 0.91

Overall correctness: patch is correct
Overall confidence: 0.82

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against a1063aa4c83f.

Label changes:

• add P2: This fixes a real provider/auth runtime bug with limited blast radius and no evidence of an active production-wide outage.
• add merge-risk: 🚨 auth-provider: The PR changes runtime model resolution based on auth profiles and ClawRouter credentials, where incorrect scoping could affect provider routing or credential behavior.
• add rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦪 silver shellfish and patch quality is 🐚 platinum hermit.
• add status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: The PR body lists targeted tests, typechecks, lint, and docs checks, but no redacted terminal output, logs, screenshot, recording, or artifact from an actual OpenClaw agent run using a ClawRouter auth profile. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.

Label justifications:

• P2: This fixes a real provider/auth runtime bug with limited blast radius and no evidence of an active production-wide outage.
• merge-risk: 🚨 auth-provider: The PR changes runtime model resolution based on auth profiles and ClawRouter credentials, where incorrect scoping could affect provider routing or credential behavior.
• rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦪 silver shellfish and patch quality is 🐚 platinum hermit.
• status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: The PR body lists targeted tests, typechecks, lint, and docs checks, but no redacted terminal output, logs, screenshot, recording, or artifact from an actual OpenClaw agent run using a ClawRouter auth profile. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.

PR surface:

Source +65, Tests +59, Docs +5. Total +129 across 4 files.

What I checked:

• Repository policy applied: Root, extensions, and docs AGENTS.md were read; the release-owned changelog rule and provider/auth compatibility guidance are directly relevant. (AGENTS.md:234, a1063aa4c83f)
• Current main behavior: Current main's ClawRouter catalog path resolves a profile-scoped credential for discovery, but the provider has no dynamic runtime model hooks, so runtime resolution cannot refresh auth-profile-only catalog rows. (extensions/clawrouter/index.ts:84, a1063aa4c83f)
• Runtime hook contract: The existing plugin SDK already supports async prepareDynamicModel followed by resolveDynamicModel for network-backed model resolution, which makes this a narrow provider-owned fix rather than a new core seam. (src/plugins/types.ts:501, a1063aa4c83f)
• Async caller path: resolveModelAsync calls prepareProviderDynamicModel before retrying model resolution, passing config, agent/workspace directories, provider config, and auth profile context to the provider hook. (src/agents/embedded-agent-runner/model.ts:1745, a1063aa4c83f)
• PR diff evidence: The PR adds provider-local dynamic model cache/refresh code and a focused auth-profile test, but also adds a hand-written CHANGELOG entry. (extensions/clawrouter/index.ts:113, 791103dc74ae)
• Related work: The broader ClawRouter canonical-provider routing PR remains open and draft with review blockers; it overlaps conceptually but does not supersede this narrower provider-id fix. (2d22ef440732)

Likely related people:

• steipete: Merged current-main ClawRouter routing, quota, docs, and tests in feat(providers): add ClawRouter routing and quotas #99658, which this PR directly extends. (role: introduced behavior and recent feature owner; confidence: high; commits: 4a354f76c1bf; files: extensions/clawrouter/index.ts, extensions/clawrouter/provider-catalog.ts, docs/providers/clawrouter.md)
• Dallin Romney: Blame and history on the dynamic model hook/caller surfaces point to recent work consolidating the async model resolution and plugin runtime files used by this fix. (role: recent adjacent contributor; confidence: medium; commits: c7aca4f02938; files: src/agents/embedded-agent-runner/model.ts, src/plugins/types.ts, src/plugins/provider-runtime.ts)

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.