DBA-grade migration safety: expand-contract plans, lock-behavior analysis, batched backfills, forward+rollback and integrity checks (Postgres/MySQL · Prisma/Rails/Django)
Find the inventory-sync bugs that don't throw errors — before your accountant does. Hand over 3 CSV exports (ERP stock, SKU mapping, channel catalog) and get back the exact SKUs silently mis-syncing right now, ranked by severity and explained. No API keys, no credentials — read-only by design.
Your AI built the app. Find out if it is ready to ship. Audits apps built with Cursor, Claude Code, Lovable, Bolt or v0 and ranks what will break after launch, worst first: auth holes, IDOR, Supabase RLS, exposed secrets (always masked), SQL/NoSQL/command injection, SSRF, XSS, payment webhooks and deploy blockers, plus SEO and AI-search readiness. Get a ship decision, scores, evidence-backed findings (file:line) and copy-paste fix prompts. Audit-only; your code is never executed.
Review your Merchant product CSV before submission. Get duplicate-ID, price/currency, required-field and supplied-GTIN checks, a row-linked HTML report, an editable fix queue and hashed evidence. Runs locally with Python; no paid API or account connection. Partial preflight, not Google approval.
"Rockstar" and "fast-paced" repel top candidates. Requiring a degree for a role that doesn't need one cuts 65% of the US workforce. Missing a salary range violates laws in 18 states — fines up to $250K. Paste your job post. Get a bias report, salary compliance check, repellent phrase detection, and an optimized version with a before/after score. HIRE framework: Hidden bias · Information completeness · Readability · Engagement. Every change explained.
Audit a machine learning platform against 62 production checks, or generate a verified AWS MLOps repository.
Before you install a third-party skill, scan it. Skill Security Audit audits a skill package — code and SKILL.md prose alike — across 10 risk classes: plain-text instruction poisoning, hard-coded secrets, prompt injection including delegated sub-agent calls, destructive commands, data exfiltration, download-and-execute chains, over-broad permissions, and disclosure gaps. Judged under your agent's permissions. Reports findings with file and line evidence. Never runs the code it audits.
Know if your backtest is a real edge or overfit noise. Runs bar-permutation MCPT, Deflated Sharpe, PBO/CSCV, Hansen SPA, Romano-Wolf StepM, a stationary bootstrap and an out-of-sample split on your own returns — no market feed, no API keys.