Use Cases / AI Code Review

Code review by 6 specialist agents

One command audits your last 20 commits. Security vulnerabilities, UX issues, architectural bottlenecks, and conversion gaps — all rated by severity, all in one pass.

The command

> /acepilot review

One command triggers five specialist agents across your recent commits. Each agent has a narrow focus and a clear mandate. Here's how it runs:

What AcePilot does

Git scan

Reads the last 20 commits with acepilot: prefix. Deduplicates — won't re-review what's already been reviewed in a previous session.

@reviewer — code quality

Checks for bugs, spec drift, and test coverage gaps. Reads TASKS.md to verify every change matches original intent.

@designer — UX + accessibility

Audits HTML, CSS, and JSX changes. WCAG 2.1 AA compliance, mobile-first rendering, and copy quality on every touched component.

@security — vulnerabilities

Scans for OWASP Top 10. Checks auth changes, API endpoints, and all user input handling. No false positives — findings are exploitable or they aren't filed.

@architect — architecture

Detects N+1 queries, coupling increases, missing pagination, and bundle bloat. Flags technical debt before it compounds.

@strategist — product

Checks conversion paths, analytics event coverage, and docs completeness for any new feature shipped. Nothing ships untracked.

What you get

must fix should fix nice to have

Real example

A review session on a Stripe webhook handler caught: a missing signature verification (exploitable immediately), no idempotency key (double-charge risk), and logging of the full payload (PII exposure). All three found in one /acepilot review. Fixed before the code hit production.

When to use this

Try it yourself

$ git clone https://github.com/acepilot-app/acepilot.git
$ cd acepilot && ./acepilot-14.0/install.sh

# In your project directory
$ claude
> /acepilot review

Audit your last 20 commits now

Free tier. No credit card. Install in 30 seconds.

Install AcePilot free