init-review-policy

Initialize Review Policy

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "init-review-policy" with this command: npx skills add ajoslin/dot/ajoslin-dot-init-review-policy

Initialize Review Policy

Create a repo-local review policy package for use by /code-review and /review .

Goal

Initialize these files:

  • .opencode/review/policy.md (required)

  • .opencode/review/checklist.md (optional but recommended)

  • .opencode/review/severity.yml (optional but recommended)

Behavior

  • Detect repo root (prefer vcs-detect if available).

  • Create .opencode/review/ if missing.

  • If files do not exist, create them from templates below.

  • If files exist, preserve user content and only add missing sections.

  • Ask for domain-specific overrides from user input and apply them.

  • Return a short summary with created/updated file paths.

policy.md template

Review Policy

Scope

  • Applies to all code reviews in this repository.
  • Overrides generic review defaults where explicitly stated.

Critical Domains

  • Authentication and authorization
  • Data integrity and migrations
  • Secrets, credentials, and PII handling
  • Billing, quota, and financial calculations

Must-Flag Findings

  • Security vulnerabilities with practical exploit paths
  • Silent data loss or corruption risks
  • Backward-incompatible API or schema changes without migration plan
  • Missing rollback/guardrails for risky deploy paths

Usually Ignore

  • Pure style nits unless they hide correctness issues
  • Hypothetical edge cases without realistic trigger paths

Repo-Specific Rules

  • Add project rules here (framework constraints, architecture boundaries, test expectations)

Required Review Output

  • Severity: critical | high | medium | low
  • File and line reference for every issue
  • Why this is a bug/risk in this repository
  • Concrete fix suggestion

checklist.md template

Review Checklist

Correctness

  • Logic matches intended behavior and existing contracts
  • Error handling is explicit and testable

Security

  • No new injection/authz/secrets/PII exposure paths

Data and Migrations

  • Schema changes include compatibility and rollback notes

Performance

  • No obvious unbounded hot-path regressions

Operations

  • Logging/metrics/alerts are sufficient for new risk areas

Testing

  • Critical paths have adequate coverage for changed behavior

severity.yml template

severity: critical: - remote code execution - auth bypass - irreversible data loss high: - privilege escalation - data corruption risk - breaking migration without rollback medium: - reliability regression on common paths - significant performance regression low: - minor maintainability risk - non-blocking robustness gaps rules: require_file_line_reference: true require_concrete_fix: true deduplicate_findings: true

Notes

  • Keep rules concise and specific to this repository.

  • Prefer concrete examples over abstract policy language.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

session-export

No summary provided by upstream source.

Repository SourceNeeds Review
General

find-skills

No summary provided by upstream source.

Repository SourceNeeds Review
General

kimaki-tools

No summary provided by upstream source.

Repository SourceNeeds Review
General

skill-creator

No summary provided by upstream source.

Repository SourceNeeds Review