Plan a feature into a design spec and linked GitHub issue queue.
npx skills add https://github.com/berlinguyinca/autospec --skill autospec-define
Please help me install this Agent Skill. Skill: autospec-define Source: https://github.com/berlinguyinca/autospec/tree/main/skills/autospec-define Command: npx skills add https://github.com/berlinguyinca/autospec --skill autospec-define
Turns a feature idea into a durable, reviewable design spec and a linked GitHub issue tree that an implementation agent can execute. It reduces the risk of vague requirements by forcing structured architecture, API, data model, errors, and testing into a spec before code changes begin. It also ensures work is explainable after the fact by stopping at the planning phase and handing off to implementation.
Invoke autospec-define with your feature request to generate a design spec in docs/specs and a linked GitHub issue queue, then continue by running /autospec-run --profile <name>.
High-intent search queries and answers about installing and using this skill.
Decompose a design spec by selecting a tracked docs/specs/*.md file, skipping research and design phases, and splitting it into a structured GitHub issue queue containing an epic plus sized child issues ready for implementation handoff.
Plan a feature through a Phase 0–3 pipeline that bootstraps the repo, investigates the codebase, produces a structured design spec covering architecture and data models, and decomposes it into linked GitHub issues before implementation begins.
Yes, you need git, gh CLI, curl, jq, and bash installed to bootstrap repositories, create GitHub issues, query API responses, and execute the planning pipeline that generates design specs and linked issue trees.
Yes, existing spec mode selects a tracked docs/specs/*.md design document, skips research and design phases entirely, and decomposes the already-written spec into the same epic plus child issue queue structure for implementation handoff.
The pipeline stops after Phase 3 to ensure work remains reviewable and explainable, handing off execution to /autospec-run for autonomous implementation. YAML frontmatter metadata is required to track spec availability and structure the design doc.
After Phase 3 completes, continue by running /autospec-run with a profile name to autonomously implement the planned issues, create pull requests, and execute the structured design spec without manual intervention.
pradeepmouliGenerate baseline-spec.md and current-state.md from repository analysis.
berlinguyincaApply Phase 3.5 ctx and reasoning labels to GitHub auto-implement issues.
berlinguyincaSplit tracked design specs into GitHub issue epics and child queues.
robertzengcnConvert vague product intent into executable specification documents and GitHub issues.
