Hi Product Hunt 👋 I'm Hetul, a CS grad student, and I built Aegisure solo. The problem: my team kept adding AI coding agents — Claude Code, Cursor, Codex, Copilot — to the same repos, and there was no consistent way to set rules across all of them or to see what each agent actually changed. Every agent has its own rule format, and once a commit lands you can't easily answer: who approved this, what rules did it follow, did it touch secrets, did it skip tests? Aegisure is two parts, both free during beta: - Free, open-source CLI: write one rule file ("constitution") and it exports to every agent's native format. It scans diffs for secrets, auth/payment changes, and skipped tests before you push — no LLM, runs offline. - GitHub App + dashboard: scans PRs, posts an advisory verdict + a check, keeps an audit trail. It advises — a human still merges. It's an early beta and I care more about honest feedback than upvotes. If you run AI agents on your repos: what risky patterns would you actually want flagged? And where does this fall short? Thanks for taking a look. 🙏

Aegisure

Control & audit layer for your AI coding agents

Control & audit layer for your AI coding agents