About
Your AI pair programmer flatters you. Sally doesn't.
Cynical Sally is the senior engineer your code hoped it'd never meet: brutally honest code reviews right in your terminal and inside your AI editor. She scores your code 0 to 10, backs every issue with evidence, and hands you fixes you can actually use. Because "You're absolutely right" is probably absolutely wrong.
Runs two ways: a CLI (npm i -g @cynicalsally/cli) and an MCP server inside Claude Code, Cursor & Windsurf: so she fits the workflow you already have.
Six tools, one honest reviewer:
Roast — code review with a score, real issues, and actionable fixes. Auto-detects what to review: staged changes, your last commit, or the whole directory.
Explain — turns the spaghetti someone left behind into plain English.
Refactor — before/after, side by side, with the reason it matters.
PR Review — reviews your PR like a senior engineer with opinions and no reason to be polite.
Frontend Review — finds why your component re-renders on every keystroke and why your z-index is load-bearing.
Brainstorm & Marketing — pressure-tests your architecture ideas and your copy before your users do.
What makes it different:
Privacy you can verify, not just trust. Run `sally roast --dry-run` and Sally prints the exact payload — every file, byte size, token estimate, and a SHA-256 receipt — then sends nothing. Secret files (.env, keys, certs) are skipped on your machine before anything leaves it.
Your code is never stored, logged, trained on, or sold. Processed in memory, then discarded.
Built for real workflows: CI/CD ready (--fail-under, --ci), JSON output, background mode with OS notifications, and opt-in share links that publish only the score — never your code.
Free to use: 90 free roasts per month, no account needed. Every premium tool includes a free trial.