TejasSinha12/Astraeus — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me an AI workspace called Astraeus where a user can give a coding mission and watch a team of AI agents plan, build, critique, optimize, and audit the result in real time. I want a clean web dashboard with areas for the main workspace, an arena for comparing different attempts, an archive of past missions, a Forge view that runs a few competing approaches side by side, and a Chronos replay view where I can scrub through the reasoning and code changes step by step.

Please include a backend API that coordinates the swarm, streams progress to the frontend, saves missions, branches, traces, agent scores, and usage data, and supports basic roles, scoped API keys, rate limits, and billing credit tracking. It should be deployable with Docker for the backend and Vercel for the frontend. Keep the UI polished but focused on researchers and teams who want to inspect how the AI made decisions. Look up current docs online if you need to.

Want more depth? Deep Reverse