Stage whole skill dirs (companions resolve) + three skill-text fixes

[aarontrowbridge]

Closes #393.

What

• stageOpencodeSkills now copies each resolved skill's whole dir (fs.cpSync recursive) instead of SKILL.md alone — companion files (tdd's 5 craft docs, 9 references/ dirs, …) land beside the staged SKILL.md so its relative links resolve. No guard change: the stage root still holds only the resolved set, and skillGrants already granted the entire source skill dir, so nothing new becomes readable.
• improve-codebase-architecture: engine-neutral explore dispatch (names Task/Agent per engine).
• amico-slack: example link repointed from the retired amico-plugin to amicode.
• amico-lab: device-lock instruction creates the parent dir.

Verification

• vitest run test/scores/package_skills.test.ts — 28/28 (incl. new companion-staging test).
• tsc --noEmit — clean.
• Full non-slow suite: 913 passed, 2 failed — both failures pre-exist on pristine main (verified via stash on the base commit): golden_parity.test.ts byte-mismatches from the WS1 — The pack interface: one manifest per domain, score as a field of a pack #391 pack merge. Out of scope here; flagging for the pack work.

[coderabbitai]

Warning

Review limit reached

@aarontrowbridge, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 16 minutes

Limit details: You’ve used all 3 included reviews currently available under your plan.

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 Plus

Run ID: e9d0e51b-8899-4398-bd6c-48d360152b09

📥 Commits

Reviewing files that changed from the base of the PR and between 84a0784 and 1085ad6.

📒 Files selected for processing (5)

• packages/extension/skills/amico-lab/SKILL.md
• packages/extension/skills/amico-slack/SKILL.md
• packages/extension/skills/improve-codebase-architecture/SKILL.md
• packages/extension/src/scores/package_skills.ts
• packages/extension/test/scores/package_skills.test.ts

---

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

[aarontrowbridge]

Rebased onto main (1085ad6) to pick up the path-portable goldens from #392 — the fast failure was the stale golden_parity base, as flagged. CI re-running; golden parity + config suites verified green locally post-merge.