feat: Enrich peek_bundle + lock to expose bundle mcp inputs and secrets

[mabry1985]

Summary

Enriches the archetype-preview (peek_bundle) response and the bundle lock to expose the inputs an archetype will ask the operator for — pure data enrichment, no behavior change.

• ops/plugins.py — _peek_bundle_sync now surfaces the bundle manifest's mcp (catalog-shaped {template, inputs: [{key, label, …}]} items) and secrets ({key, label, placeholder, secret, required} items) in the peek result, so ArchetypePreviewDialog can show what the archetype will collect without installing. The single-plugin (non-bundle) peek branch carries both keys as [], so consumers read mcp/secrets uniformly across bundle and plugin previews.
• graph/plugins/installer.py — _install_bundle caches secrets in the plugins.lock bundle entry, mirroring the existing mcp cache, so the lock-only workspace-create path can seed/prompt for these inputs without re-parsing the bundle.
• Tests — test_ops_plugins.py gains coverage that a bundle peek surfaces mcp/secrets and that a single-plugin peek returns both empty; test_plugin_installer.py asserts the lock now caches mcp and secrets alongside the existing provenance fields (shared fixtures updated to declare both fields).

Refs #2041 (slice 1 of 3)

[coderabbitai]

Warning

Review limit reached

@mabry1985, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 31 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: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: c0894e0c-8b0c-431b-9add-1fe974f23e38

📥 Commits

Reviewing files that changed from the base of the PR and between b080884 and ad5db4b.

📒 Files selected for processing (4)

• graph/plugins/installer.py
• ops/plugins.py
• tests/test_ops_plugins.py
• tests/test_plugin_installer.py

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/bd-3pf

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[protoreview]

QA panel review — PASS

code-review-structural · head ad5db4b40db0 · formal

[review-synthesizer completed: workflow code-review-structural:report]

Prose Brief

Overall risk: Very low. The PR is 100% additive — every hunk is a + line, no existing behavior is altered, and the list(bundle.get("X") or []) guard pattern is identical to the one already proven in the "mcp" handling within installer.py. Both the bundle and single-plugin peek paths and the install lock path are covered by new tests.

Fix-first item: Nothing to fix. All four LLM finders converged on clean, and the verifier confirmed every claim against the diff and default-branch code.

Panel disagreement: None — unanimous clean verdict.

What verification changed: No findings were refuted or flagged uncertain. Every claim was SUPPORTED. Verification did surface one factual correction: the synthesis brief stated "mcp" was "plumbed" into the install lock path, but "mcp" already existed there in the default branch — only "secrets" is new to the lock; "mcp" is new only to the peek path. This doesn't affect any finding.

Gaps: The structural lane (protoPatch) errored with a rate-limit and produced no findings — a tooling gap, not a code defect. Both modified production files (ops/plugins.py, plugins/installer.py) were covered by the LLM panel, so coverage is adequate despite the structural miss.

---

[]

[protoreview]

Promoting the PASS verdict for head ad5db4b40db0: all checks terminal-green, zero unresolved review threads. (approve-on-green)