audit

name	audit
description	Full quality gate — architecture, tests, docs, and build verification
user_invocable	true

/audit

Comprehensive 5-pass quality gate for Alder Grove. Runs all sub-audits and reports overall PASS/FAIL.

Arguments

$ARGUMENTS controls which sub-audits to run:

• /audit — run all 5 passes
• /audit architecture — run only Architecture pass
• /audit docs tokens — run only Documentation and Design Tokens passes
• /audit build security — run only Build & Test and Security passes

Valid names: architecture, build, docs, tokens, security

If $ARGUMENTS is empty or not provided, run all 5 passes.

Passes

Pass 1: Architecture Compliance

Run the /check-architecture skill (reads .claude/skills/check-architecture/SKILL.md).

This dispatches the frontend and backend architecture specialists and aggregates results.

Pass 2: Build & Test

Run these commands sequentially — all must pass:

# Frontend
pnpm check          # TypeScript type checking + ESLint
pnpm test           # Vitest unit tests

# Rust
cargo build --workspace       # Build all crates
cargo test --workspace        # Run all Rust tests
cargo clippy --workspace -- -D warnings   # Lint Rust code (warnings = errors)

Any failure in this pass = overall FAIL.

Pass 3: Documentation Consistency

Run the /audit-docs skill (reads .claude/skills/audit-docs/SKILL.md).

If the skill file does not exist, report as SKIP with WARNING.

Pass 4: Design Tokens

Run the /audit-tokens skill (reads .claude/skills/audit-tokens/SKILL.md).

If the skill file does not exist, report as SKIP with WARNING.

Pass 5: Security

Run the /audit-security skill (reads .claude/skills/audit-security/SKILL.md).

If the skill file does not exist, report as SKIP with WARNING.

Parallelization

Passes 3, 4, and 5 are independent of each other. When running all passes, dispatch them via the Agent tool in parallel where possible. Passes 1 and 2 may also run in parallel with each other but their results must be collected before the final summary.

Output Format

=== ALDER GROVE AUDIT ===

| #  | Section              | Status | Findings | Top Issue                        |
|----|----------------------|--------|----------|----------------------------------|
| 1  | Architecture         | PASS   | 0        | —                                |
| 2  | Build & Test         | PASS   | 0        | —                                |
| 3  | Documentation        | PASS   | 1        | /audit-tokens missing from table |
| 4  | Design Tokens        | FAIL   | 3        | Hardcoded #ff0000 in Panel.tsx   |
| 5  | Security             | SKIP   | —        | Skill not found (WARNING)        |

=== OVERALL: FAIL — 3/4 PASS (1 SKIP) — 4 total findings ===

Rules

• Run ALL requested passes — do not skip any unless the skill file is missing
• A pass is FAIL if it has any HIGH or CRITICAL findings
• A pass is PASS if it has only LOW or MEDIUM findings (or none)
• A pass is SKIP if its skill file does not exist (shown as WARNING)
• Overall result: FAIL if ANY pass is FAIL; PASS if all passes are PASS or SKIP
• Skipped sections count toward neither PASS nor FAIL in the denominator
• Report the total number of passes that were PASS out of those actually run
• When $ARGUMENTS limits scope, only count the requested passes
• Run this before creating PRs or claiming work is complete