Skip to content

AL-JANEF/janef-forge / skills/engineering/review-defect-first

review-defect-first

Verified

Independent review of a diff, PR, or file that hunts concrete defects introduced by the change: correctness, security, data loss, authorization or tenant escape, production failure, broken contracts, performance regressions, missing critical tests. Severity-ordered findings citing a real code path; no style inflation; "no qualifying defect" is valid. Use for code review and as the independent gate at high risk.

FileKindSize
SKILL.mdskill md2.1 KB
capability.jsonconfig1.9 KB
references/review-checklist.mdmarkdown2.2 KB

references/review-checklist.md

View on GitHub
# Review Checklist

Walk every dimension that the diff touches. A gap in any is a finding, not an oversight.

**Correctness** — Does it do what was asked? Edge cases, boundary values, empty and error inputs handled? Off-by-one, null/undefined, timezone, encoding traps checked?

**Architecture** — Does it fit the existing design or fight it? Are boundaries and responsibilities respected? Any dependency-direction or ownership violation introduced?

**Security** — Red-team it. Injection (SQL/NoSQL/command), XSS, CSRF, SSRF, IDOR, privilege escalation, path traversal, secrets exposure, broken authentication or authorization. In multi-tenant code: is every query tenant-scoped? Deep remediation: `vuln-audit`, `auth-hardening`, `secrets-guard`.

**Data** — Migrations reversible or forward-fixable? Constraints preserved? Transactions around multi-step writes? Backfills batched?

**Performance** — N+1 queries, unindexed lookups, memory growth, blocking calls on hot paths, concurrency races and deadlocks.

**Error handling** — Failures caught and handled meaningfully, not swallowed. No empty catch blocks. Errors surface actionable information without leaking internals.

**Logging and observability** — Security-relevant and diagnostically important events logged without secrets or PII (`security-logging`). Can production behavior be reconstructed from what is emitted?

**Tests** — Do new tests assert behavior that matters? Could they fail? Are critical paths (auth, money, data integrity, tenant boundaries) covered by a denied case and a failure case?

**Maintainability and readability** — Honest names, cohesion, no dead code, no duplication, no commented-out blocks, no cleverness that obscures intent.

## Forbidden-pattern scan

Reject on sight: `TODO`/`FIXME` presented as complete · placeholder or stub business logic · mock logic for real behavior · dead code · copy-paste duplication · disabled tests · disabled lint or type checks · `@ts-ignore` · unjustified `any` · hard-coded secrets.

## Acceptable evidence for a finding

A file and line range plus the call site, invariant, or test gap that makes it a defect. Describe the impact; do not write a working exploit.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall