feat(agent-attention): automate approval blocker lifecycle

[nathanvale]

Summary

Agent Attention V1 could deliver one completed reminder to its exact Codex task. This PR makes that approval lifecycle discoverable from any task and closes the terminal receipt path.

A global route and thin skill submit structured yes/no unblockers to one owner. The owner validates exact meaning and task binding, creates one native alert-backed gate, suppresses duplicate submissions, delivers completion once, and appends one bounded terminal outcome to the retained reminder. A Stop hook observes structured owner state only; it never infers approval from assistant prose or transcripts.

The persistent 15-second background target is not qualified by this PR. The local watcher meets the target, but the supported Codex scheduled-task owner runs at minute granularity. No live Apple Reminder was created or changed during milestone-two implementation.

Design decisions

• Admit only explicit yes/no decisions that unblock a paused task.
• Treat no response as paused work. Never auto-approve or nag.
• Keep policy in the Python owner; keep the skill and hook as thin adapters.
• Bind reminders, deliveries, and outcomes to exact task and stable reminder IDs.
• Use exclusive claims and append-only receipts for duplicate suppression and unknown-result recovery.
• Keep disagreement and multi-choice decisions in the linked Codex task.

Validation

• bun run test:agent-attention: 17 Python owner tests and 6 Stop-hook tests passed.
• Local fake-EventKit harness: 0.003 seconds from completion to detection and 0.046 seconds to recorded delivery receipt; duplicate notification count was zero.
• bun run biome:check: passed across 941 files.
• Python compile, tab checks, and git diff --check: passed.
• Real remindctl help confirms the used --list-id, --alarm, --url, --priority none, show all, and show completed surfaces.
• Durable project vault check: passed.
• Full bun test: 6,529 passed, 2 skipped, 37 failed, and 6 errors. The failures are outside this diff and include existing instruction-owner, fixture, setup-projection, and archived-package failures.

Residual qualification

• Decide whether the product keeps the 15-second background target or accepts the supported one-minute owner.
• After that decision, obtain fresh exact approval for one dedicated milestone-two live Reminders smoke gate.

[changeset-bot]

⚠️ No Changeset found

Latest commit: 04d7178

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in: 40 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: db95e730-6be0-4287-94a2-ce56d91b6c06

📥 Commits

Reviewing files that changed from the base of the PR and between 9256872 and 04d7178.

📒 Files selected for processing (9)

• .codex/hooks.json
• AGENTS.md
• hooks/agent-attention-codex-stop.test.ts
• hooks/agent-attention-codex-stop.ts
• package.json
• runtime/agent-attention/README.md
• runtime/agent-attention/agent-attention.py
• runtime/agent-attention/test_agent_attention.py
• skills/agent-attention/SKILL.md

---

_{Comment @coderabbitai help to get the list of available commands.}

[nathanvale]

@coderabbitai review

[coderabbitai]

⚠️ Action not completed

Review rate limited.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[nathanvale]

Superseded by the installed plugin implementation merged through myagentdojo/agent-plugin-playground#3. The plugin path passed exact-head review and CI, installed-path automation, fresh Codex discovery, and one approved live Agent Attention gate. Closing this PR without deleting its migration-evidence branch.