[freeqaz-openai]

Agent detail now keeps deployment activity, current diagnostics, and failure recovery in separate TypeScript modules. They use the controller’s existing response types and ship at the same Console URLs. A late first status read still fills the space reserved above the version tabs. Polling, recovery permissions, and the distinction between a recorded result and live serving stay the same.

The 57 deployment browser checks passed, including late status, a failed read followed by recovery, and desktop and phone layout. In the first parallel run one check timed out waiting for a completed record. It passed by itself, then the full set passed serially without changing the test. Strict Console TypeScript, changed-file lint and formatting, the public asset checks, and the docs and Storybook builds passed.

In Pages / Agent detail / New version failed and Current observations for a reader without operate, before and after screenshots matched at 1280 and 390 pixels wide. The walkthrough opens failure details, follows the affected version’s Logs, and shows diagnostics unavailable to a reader. These are simulated Storybook states.

Storybook evidence

deployment-evidence-and-recovery.webm

[clawsweeper]

🦞👀
ClawSweeper picked this up.

Pull request received. I will update this pull request when review starts.

ClawSweeper review complete

ClawSweeper finished reviewing this revision. The review result is being finalized.

View the workflow run.

[clawsweeper]

Codex review: needs maintainer review before merge.

What this changes

This PR extracts Agent deployment evidence, diagnostics, and recovery into typed Console modules while preserving existing behavior.

Example: Deployment v7 fails while v6 remains selected.

• Before: The Agent detail module renders the failure evidence and permitted recovery links.
• After: Typed modules render the same evidence and permitted recovery links.

Review scores

Measure	Result	What it means
Overall readiness	🦐 gold shrimp (3/6)	The extraction is coherent and has no concrete findings; behavioral evidence remains scoped to simulated UI and browser integration coverage.
Proof confidence	🦐 gold shrimp (3/6)	This maintainer-authored refactor is outside the ordinary contributor proof gate; inspected Storybook screenshots support preserved browser presentation, while the reported browser checks support polling and recovery behavior without establishing live deployment or model execution.
Patch quality	🐚 platinum hermit (4/6)	No actionable review findings were identified.

Product

Kind: Refactor · Worth it: Yes
Reason: The maintainer-authored direction separates substantial existing responsibilities and adds type checking without changing public contracts or defaults.

Merge readiness

✅ Ready for maintainer review

This PR is a useful, maintainer-authored refactor with no actionable correctness findings; keep it open for normal landing checks.

Priority: P3
Reviewed head: dc1fbef5cdadd99be2eb692709bb6f545c435515

Before merge

None.

Findings

None.

Agent review details

How this fits together

The Console reads authorized Agent, revision, deployment, and permission responses and renders deployment history, current observations, and recovery navigation.

flowchart LR
  API[Controller API] --> Detail[Agent detail]
  Detail --> Evidence[Deployment evidence]
  Evidence --> Diagnostics[Current diagnostics]
  Detail --> Recovery[Failure recovery]
  Evidence --> Browser[Browser view]
  Diagnostics --> Browser
  Recovery --> Browser

Loading

Technical review

Best possible solution:

Retain the typed extraction and existing permission, polling, and presentation contracts.

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

This PR proposes a refactor rather than a bug repair; source comparison verifies the extraction without running target code.

Is this the best way to solve the issue?

The three modules replace their original implementations and use existing response contracts, with no competing runtime path or speculative capability.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning medium; reviewed against fdb81427bc51.

Provenance checked

• Deployment overview and diagnostics keeps the original intent (7c3ecbf: Separate selected version, recorded deployment work, and live observations while retaining exact-resource authorization.)
• Failed-version recovery keeps the original intent (4bf2532: Put the affected version's recovery actions first and keep recorded completion distinct from serving health.)
• Configuration readability before recovery keeps the original intent (765417d: Avoid offering Deploy again before a successful Configuration read establishes a usable destination.)
• Late initial deployment reads keeps the original intent (58bbb8a: Keep the version tabs stable when the first deployment record arrives after the page reveal.)

Testing

Proof path: in-process harness.

Security

None.

Evidence

What I checked:

• Verified PR scope: The introduced changes move existing rendering and lifecycle logic into three typed modules, register their browser assets, update imports, and revise the owning flow documentation. (apps/controller/src/console/agents/detail.mjs, dc1fbef5cdad)
• Recovery gates preserved: Deploy again still requires an affirmative deploy permission and a successful Configuration read, preserving the behavior established in fix(console): offer Deploy again only after the Configuration read succeeds #2883. (apps/controller/src/console/agents/deployment-recovery.mts:212, dc1fbef5cdad)
• Existing product contracts retained: The extraction retains recorded-versus-live status separation from Clarify Agent deployment versions and add on-demand diagnostics #482, recovery presentation from Clarify Agent status and put failed-version recovery in the first screen #2542, and late-read disclosure behavior from fix(console): keep the version tabs still when the first deployment read is late #3160. (apps/controller/src/console/agents/deployment-evidence.mts:509, dc1fbef5cdad)
• Latest revision repairs navigation expectations: The follow-up adds exactly the three new browser module GETs to the navigation request inventory; existing API expectations and the no-write assertion remain intact, matching the author's discussion. (tests/browser/console.test.mjs:430, dc1fbef5cdad)
• Visual evidence inspected: All three downloaded screenshots were inspected: phone recovery for failed v7, expanded deployment evidence, and disabled diagnostics for a reader without operate access. These are explicitly simulated Storybook states. (c8fd02de7d41)
• Video inspection infrastructure limit: The downloaded recording has no prepared contact sheet because the manifest reports 'ffprobe failed: spawnSync ffprobe ENOENT'; preparing browser playback also failed because the designated scratch directory returned Errno 30, read-only file system. The recording was not visually assessed.

Labels

Label changes:

No label changes.

Label justifications:

• P3: This is an internal typing and ownership refactor with preserved user-facing behavior.
• rating: 🦐 gold shrimp: Overall readiness is 🦐 gold shrimp; proof is 🦐 gold shrimp and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR.
• proof: 📸 screenshot: Contributor real behavior proof includes screenshot evidence.

Rating scale

6/6 🦀 challenger crab · 5/6 🦞 diamond lobster · 4/6 🐚 platinum hermit · 3/6 🦐 gold shrimp · 2/6 🦪 silver shellfish · 1/6 🧂 unranked krab. Overall follows the weaker of proof and patch quality; ✨ marks media proof (a screenshot, video, or linked artifact) that directly shows the changed behavior.

Workflow

ClawSweeper edits this one comment on every review. Comment @clawsweeper re-review for a fresh review only. Repair and merge execution lanes are retired.

History

Review history (1 earlier review cycle)

• reviewed 2026-10-22T11:25:03.299Z sha c8fd02d :: needs maintainer review before merge. :: none

Reviewed October 22, 2026, 7:32 AM ET / 11:32 UTC (Revision 2).

[freeqaz-openai]

Fixed the Console Browser failure in dc1fbef. The prior run passed 475 of 476 cases; the one failing navigation test correctly observed the three new deployment module GETs, which its exact request list did not yet include. I reproduced the failure and added just those three URLs. The API request expectations and no-write assertion are unchanged. The targeted test and all 50 cases in its suite passed locally.

@clawsweeper re-review

[clawsweeper]

🦞👀
Exact review queued.

Re-review progress:

• State: Complete
• Detail: The durable review result and its route handoff completed.
• Run: https://github.com/openclaw/clawsweeper/actions/runs/39771477054
• Updated: 2026-10-22T11:33:47.973Z