corgispec-propose

Generate complete engineering planning artifact packages with dependency ordering.

113|9|Updated Apr 30, 2026
One-click install
npx skills add https://github.com/ricoyudog/Coding_Corgi_flow --skill corgispec-propose
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: corgispec-propose
Source: https://github.com/ricoyudog/Coding_Corgi_flow/tree/main/.opencode/skills/molecules/corgispec-propose
Command: npx skills add https://github.com/ricoyudog/Coding_Corgi_flow --skill corgispec-propose

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It solves the problem of turning a vague idea into a disciplined, reviewable engineering plan without losing track of artifacts, checkpoints, or downstream implementation details.

Core Features & Use Cases

  • One-step proposal packaging: Creates the full planning artifact set (proposal, specs/design, and tasks) so implementation can start immediately.
  • Deterministic artifact dependency flow: Iteratively generates required artifacts in the correct order until all applyRequires outputs are complete.
  • Optional GitLab issue tracking closeout: Creates parent + child issues (when GitLab tooling is available and tracking state doesn’t already exist) and writes a canonical .gitlab.yaml handoff contract.
  • Optional worktree isolation metadata: When enabled, creates an isolated git worktree and records .worktree.yaml so later phases operate in the right location.

Quick Start

Tell the AI to propose a change called "add user auth", and it will generate the full proposal/specs/design/tasks package ready for review and handoff.

Frequently Asked Questions about corgispec-propose

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a complete engineering proposal with specs, design, and tasks in one step?

To generate a complete engineering proposal, use a one-step workflow that derives a change name from configuration and iteratively produces specs, design, and task artifacts in the correct dependency order. This packages the full planning set for immediate review and handoff.

How does worktree isolation work when planning an engineering change?

Worktree isolation works by creating an isolated git worktree and recording its location in a `.worktree.yaml` metadata file. This ensures later implementation phases operate in the correct, separated environment without affecting the main branch.

Can I create GitLab issues automatically from an AI workflow proposal?

Yes, you can create GitLab issues automatically from a proposal when GitLab tooling is available. The workflow generates parent and child issues, then writes a canonical `.gitlab.yaml` handoff contract to close the tracking loop if state does not already exist.

What is the best way to ensure planning artifacts are generated in the correct dependency order?

The best way to ensure correct artifact generation order is to use a deterministic dependency flow that iteratively checks `applyRequires` outputs. This mechanism produces required specs, design, and task documents sequentially until all outputs are complete.

Do I need a corgispec/openespec configuration to generate task planning artifacts?

Yes, you need to read corgispec/openespec configuration to generate task planning artifacts. The workflow requires this configuration to derive the change name and enforce constraints that prevent embedding context or rules blocks into the final outputs.

What constraints are enforced when generating specs and design artifacts?

The generation process enforces constraints that strictly prevent embedding context or rules blocks into the final specs and design outputs. This ensures the planning artifacts remain clean, reviewable, and free of extraneous pipeline instructions.