Best possible solution:

Land the narrow type-cycle fix after maintainer handling if the exact head keeps the import-cycle guards, formatting, lint, and focused tooling tests green.

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

Yes. Source inspection shows the stricter madge guard collects static import/export declarations, while the PR replaces the problematic type-only static edges with import type queries; I did not rerun checks because this review is read-only.

Is this the best way to solve the issue?

Yes. The import-query approach is the narrowest maintainable fix because it preserves the existing plugin registry/public type contract and avoids weakening the architecture guard.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against b9e193ce22ae.

Label justifications:

• P3: This is a small internal architecture/tooling cleanup with no user-facing runtime behavior change claimed.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🐚 platinum hermit and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (terminal): The PR body includes after-fix terminal-style guard output and focused tooling-test output for this internal source-level change.
• proof: sufficient: Contributor real behavior proof is sufficient. The PR body includes after-fix terminal-style guard output and focused tooling-test output for this internal source-level change.

PR surface:

Source +3, Tests +1. Total +4 across 4 files.

What I checked:

• PR diff remains useful on current main: Comparing the PR head to origin/main shows the Android test expectation is already on current main, while the three plugin type-reference changes remain unique to this PR. (src/plugins/registry-types.ts:33, b59dd5e9fb4e)
• Madge guard target: The madge-style guard records static import/export declarations as graph edges, so replacing selected type imports with import type queries removes those edges from that guard without adding runtime code. (scripts/check-madge-import-cycles.ts:31, b59dd5e9fb4e)
• Runtime guard distinction: The runtime import-cycle guard already excludes type-only import/export declarations, so this PR targets the stricter static architecture guard rather than runtime value imports. (scripts/check-import-cycles.ts:74, b59dd5e9fb4e)
• Plugin registry contract unchanged: The registry shape and exported type names remain the same; only local type dependencies on the plugin type barrel and channel types move to import type queries. (src/plugins/registry-types.ts:33, b59dd5e9fb4e)
• Test routing context: Android version tests import the shared temp-dir helper through android-version.test-support.ts, matching the routed test expectation included in the branch and already present on current main. (test/scripts/android-version.test-support.ts:5, b59dd5e9fb4e)
• Live PR state: GitHub reports this PR open, non-draft, cleanly mergeable, labeled maintainer, and with successful relevant CI/status checks for the reviewed head. (b59dd5e9fb4e)

Likely related people:

• Vincent Koc: Recent history includes the registry type surface split and follow-up cycle/runtime registry seam work that this PR adjusts. (role: recent area contributor; confidence: high; commits: c87994bc9abd, 74e7b8d47b18, 543c14a4edd1; files: src/plugins/registry-types.ts, src/plugins/types.ts, src/plugins/runtime-state.ts)
• Peter Steinberger: Recent plugin/channel refactors and build export surface repairs touch the same plugin boundary area that the registry type surface depends on. (role: adjacent owner; confidence: medium; commits: 3e013342de26, d5cb8cebcd83, ab96520bbad5; files: src/plugins/types.ts, src/plugins/registry-types.ts, src/channels/plugins/types.plugin.ts)
• joshavant: The Android pinned release versioning commit introduced the Android version tests and shared test support that explain the test-project routing expectation. (role: adjacent test owner; confidence: medium; commits: 40eec48caf29; files: test/scripts/android-version.test.ts, test/scripts/android-pin-version.test.ts, test/scripts/android-version.test-support.ts)

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.