lfg

Autonomous engineering workflows are automated by coordinating end-to-end planning, implementation, and review steps. This reduces manual orchestration overhead and accelerates software delivery by executing a predefined pipeline across plan, work, review, and merge stages.

Automated code review and version control integration work by executing host-provided skills sequentially within the pipeline. The system enforces sequential gating, persists autofixes, handles residual review findings, and pushes pull requests automatically.

A written plan is required before starting automated feature implementation. An automated planning gate validates the existence of a plan before proceeding to implementation, ensuring structured execution of multi-step software engineering workflows.

You can autonomously plan, implement, and push a pull request by running the pipeline. It executes multi-step automation across planning, code review, and CI tool integration to handle the complete feature rollout and merge process.

Residual review findings during autonomous code generation are handled according to predefined policy. The orchestration pipeline persists autofixes and manages residual findings autonomously, reducing manual remediation overhead.

Running lfg autonomously plans, implements, reviews, and merges a feature end-to-end. It automates interaction with host-provided skills for planning, work execution, code review, and commit push operations.