OpenExec Documentation
Welcome to OpenExec — an AI delivery system that turns GitHub issues into reviewed, verified pull requests.
The bottleneck in AI-assisted development was never coding speed; it's keeping the team in sync — copy-pasting a ticket to an agent, making the change, then updating Jira again by hand. OpenExec was built on greenfield-grade precision — extract intent, blueprint the work, decompose it into tasks, and verify every result against the goal — and points that same rigor at the individual tasks your team already tracks. Label an issue and it triages, plans, builds, and verifies it unattended, then opens a PR for you to review. The trail from issue → task → PR → merge is recorded automatically. Governance is optional.
OpenExec ships as a single binary that can run the whole lifecycle — from a single task to an entire greenfield project.
Core Documentation
- The Architectural Roadmap — Foundational specifications and the v1 strategy.
- Why OpenExec? — Compare OpenExec to other AI tools and learn about Hybrid Brains.
- Getting Started — Install and initialize your first project.
- Monitoring Status — How to use
openexec statusto monitor active runs. - Blueprint Engine — How deterministic and agentic stages work together.
- State Management — SQLite as the canonical source of truth.
- Deterministic Control Plane — Learn about surgical symbols and environment topologies.
- Guided Wizard — How the AI Architect gathers requirements.
Component Deprecation
The standalone OpenExec Planner (Python package) is now deprecated and its logic has been fully integrated into the core openexec binary. New users should use the unified binary.