impl-spec-executor

Automating implementation spec execution involves creating a topic branch, spawning per-phase subagents for implementation, and enforcing gated code and security reviews. This workflow requires a ready impl-spec file, a linked design spec, and a clean working directory to orchestrate end-to-end development.

A topic branch workflow for spec execution isolates feature development by creating a dedicated branch, then systematically processing implementation phases through subagents. It coordinates phased checkpoints and review gates to ensure code quality before merging automated software engineering tasks.

Setting up automated spec execution requires an impl-spec file with status ready, a linked design spec, and a clean working directory. You also need a task system configured to track per-phase subagents and manage blockers surfaced during gated code and security reviews.

Automated code review handles blockers by enforcing review gates at each phase checkpoint using subagents. When auto-resolution fails during the implementation spec execution, the system surfaces unresolved issues to users for guidance rather than proceeding with the workflow.

Yes, you can run security reviews automatically within a phased development pipeline by spawning reviewer subagents at each phase checkpoint. The automated workflow enforces these security review gates alongside code reviews, halting execution and surfacing blockers if auto-resolution fails.

You should avoid using automated subagents for spec execution when your working directory is not clean, when your impl-spec file lacks a ready status, or when no linked design spec exists. The workflow requires these exact prerequisites to function properly.