Personal project · AI engineeringIn progress
Runveil
- Purpose
- Build the infrastructure to run AI agents with controlled tools, inspectable execution history and reproducible evaluation.
- My contribution
- Built the core agent runtime, PostgreSQL persistence, model integrations, controlled repository tools and an offline worker with checkpoint recovery.
- Planned outcome
- A software-engineering agent that investigates repository tasks, makes isolated changes, runs tests and presents patches for human approval.
- In progress
- Developing execution limits and recovery, with human approvals, a trace interface and evaluation planned.