omg-autopilot

To turn a product idea into working code autonomously, this Skill runs an end-to-end lifecycle spanning requirements expansion, technical planning, parallel implementation, and QA cycles. It converts a brief input into a validated implementation without requiring manual intervention for each phase.

Autonomous code generation executes multi-phase software delivery by expanding vague product requirements into structured specs and plans. When input ambiguity is detected, it pauses execution and prompts the user with interactive decision points to clarify specifications before proceeding with implementation.

Automating the software delivery lifecycle involves orchestrating Phase 0–5 pipelines that handle requirements expansion, technical planning, execution, QA, and multi-perspective validation. You initiate this by requesting a full autopilot execution for your product idea, and the system manages the downstream phases.

Yes, autonomous execution enforces strict stop conditions after repeated QA or validation failures. If the system encounters iterative failure recovery issues or multi-perspective validation rejections, it halts the pipeline and triggers interactive hooks to request user guidance before retrying.

Yes, you can reuse generated specs and technical plans because the system writes persistent artifact outputs like spec.md and autopilot-impl.md. These artifacts are stored in designated directories, allowing you to reference and reuse them for iterative implementation rework or future product development cycles.

Autonomous implementation pauses during technical planning to confirm specifications when the generated spec needs validation or when QA is stuck. This interactive hook protocol ensures that requirements analysis aligns with user intent before committing to parallel implementation and code generation.