Infrastructure and UI building blocks on Autopilot

/loop 30m Start the "Infrastructure and UI building blocks on Autopilot" loop.

Inspired by 21st (https://21st.dev).

Goal: open work triaged, exceptions owned, and core design workflow current with audit trail
Max iterations: 20
Between iterations run: Report open queue items, stale tasks, failed automations, and items awaiting human approval for 21st
Exit when: zero open items without owner or explicit escalation, all external actions approved or sent, and systems of record current

Step 1 — Scaffold integration: Wire SDK, MCP, or agent runtime into the target environment.
Step 2 — Configure policies: Set auth, scopes, retries, and observability hooks.
Step 3 — Run smoke tests: Validate happy path and failure modes on sample workloads.
Step 4 — Harden production: Add rate limits, secret handling, and drift detection.
Step 5 — Monitor and repair: Track reliability; patch breakages when upstream APIs change.

## Before you start

Connect plugins:
- GitHub (required) — Read branches, PRs, reviews, checks, workflow runs, and source diffs.
- Figma (required) — Inspect designs, dimensions, assets, and handoff notes.
- Browser / Web access (required) — Open pages, inspect live state, collect evidence, and verify changes.

Attach skills:
- Loop runner (required) — Self-pace iterations, run the check between passes, and stop only on the exit condition.
- Code change + local verification (optional) — Edit code safely, run commands, and keep changes scoped.
- CI debugging (optional) — Read failing checks, logs, and the smallest actionable root cause.
- Approval workflows (optional) — Keep outbound actions in draft or approval states when risk is non-trivial.
- Test repair (optional) — Run tests, triage failures, and avoid weakening the suite.
- Visual design QA (optional) — Compare implementation against visual expectations and capture regressions.
- Browser QA (optional) — Exercise product flows, capture visual evidence, and verify fixes in-browser.
- Security triage (optional) — Assess alerts, prioritize patches, and avoid unsafe shortcuts.

Self-pace this loop. After each iteration, run the check command, read the output, and only continue if the exit condition is not met. Stop when the exit condition passes or max iterations is reached. Give a short status update each pass.