conch-review-health

name	conch-review-health
description	Produce a fast multi-dimension health snapshot for the Conch project and identify the highest-priority hotspots. Use when the user wants an overall review, quick risk scan, system health check, or guidance on what subsystem to inspect next.

Conch Review Health

Use this skill when the user wants a broad status view of Conch plus a clear sense of what deserves attention first, not a deep audit of one narrow issue.

Inputs

• target_repo: path to the Conch repository
• optional scope: subsystem or directory to emphasize

Default target repo: /home/stopire/Conch

Load These Assets

• policies/severity.md
• policies/report-template.md
• reviewers/lifecycle-reviewer.md
• reviewers/image-snapshot-reviewer.md
• reviewers/resource-leak-reviewer.md
• reviewers/concurrency-reviewer.md
• reviewers/api-sdk-parity-reviewer.md
• reviewers/security-boundary-reviewer.md
• reviewers/docs-implementation-auditor.md

Use these scripts:

• scripts/discover_conch.py
• scripts/collect_entrypoints.py
• scripts/normalize_findings.py when you need a unified findings envelope

Workflow

1. Discover the repo shape

Run:

python3 scripts/discover_conch.py <target_repo>

Stop if the repo does not look like Conch or if the result is too incomplete to support a reliable health snapshot.

2. Collect likely entrypoints

Run:

python3 scripts/collect_entrypoints.py <target_repo>

Use this output to identify the highest-signal surfaces for lifecycle, image/snapshot, and SDK-facing review.

3. Dispatch summary-oriented reviewers

Use the loaded reviewer specs in summary mode. Do not let any one reviewer turn this into a deep audit.

Target review dimensions:

• lifecycle soundness
• image/snapshot artifact integrity
• cleanup and residue risk
• concurrency and teardown overlap
• API/CLI/SDK parity
• security boundary posture
• docs versus implementation drift

4. Synthesize a dashboard

Use policies/report-template.md as the output skeleton.

Produce:

• repository profile
• per-dimension status: green, yellow, or red
• 1-3 top findings per dimension at most
• top 3 overall priorities
• ranked hotspots with short rationale
• recommended next workflow: conch-review-deep-review or conch-review-smoke

Constraints

• Prefer breadth over depth.
• Every yellow or red dimension should cite evidence.
• Merge duplicate findings when several reviewers are describing the same operational problem.
• Include hotspot ranking, but do not turn the skill into a full evidence-heavy deep review.
• If the user really wants one path audited in detail, switch to conch-review-deep-review.