Skip to content

unxed/f4 / .agents/skills/aif-review

aif-review

Verified

Perform code review on staged changes or a pull request. Checks for bugs, security issues, performance problems, and best practices. Use when user says "review code", "check my code", "review PR", or "is this code okay". Optional +check flag validates findings via a fresh-context subagent.

FileKindSize
SKILL.mdskill md12.9 KB
references/CHECK-MODE.mdmarkdown9.8 KB
references/SEVERITY.mdmarkdown2.7 KB
references/VALIDATOR.mdmarkdown8.9 KB

references/SEVERITY.md

View on GitHub
# Review item severity levels

This document is the single source of truth for what counts as a *critical* item versus a *suggestion* in `aif-review`. Both `SKILL.md` and `references/VALIDATOR.md` defer to this file — do not redefine the levels inline elsewhere.

## Two levels

- **critical** — merge-blocking. Lands in the "Critical Issues" section of the rendered review and contributes to `aif-gate-result.blockers`. Pick this level when the cited behavior would cause:
  - data loss, corruption, or unbounded resource usage,
  - a security regression (auth bypass, injection, leaked credentials),
  - broken correctness on a path that real users hit,
  - a downstream regression that needs a fix before merge (build break, contract change without consumers updated).

- **suggestion** — non-blocking. Lands in the "Suggestions" section and does NOT contribute to `aif-gate-result.blockers`. Pick this level when the cited behavior is:
  - a clarity / readability / naming improvement,
  - a performance-budget tweak or micro-optimization,
  - a missing log line or comment,
  - a future refactor that does not block this change,
  - a test gap that does not hide an active bug.

When in doubt between the two, leave the item at its original level. Reclassification is a deliberate judgment, not a default.

## When validator reclassifies

The `+check` validator may move an item between levels when the original review picked the wrong one. Two directions:

- **suggestion → critical** (promote) — the underlying behavior is more severe than the original framing. Example: the original review framed it as "extra database round-trip", on inspection it is "N+1 that times out the request under realistic load".
- **critical → suggestion** (demote) — the underlying behavior does not actually block merge. Example: the original review framed it as "missing null check that would crash", on inspection the upstream caller already guarantees non-null and the cited path is unreachable in production.

Reclassification with no text change is allowed (text is fine, only the level was wrong). Reclassification with a rewritten text is also allowed (both wrong).

The validator does NOT reclassify items just to balance the sections or to make the review "feel right". A move requires a concrete reason tied to the actual code or the actual cited behavior.

## Where `aif-gate-result` is computed

How `critical` / `suggestion` items project into the final `aif-gate-result` (`status`, `blockers`, `suggested_next`) is **not** defined here. The single owner of that computation is the **Machine-readable gate result** section of `SKILL.md`; `+check` reuses it via `references/CHECK-MODE.md`. This file only defines the two levels and when to move an item between them.
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