todo-dsl-generator

star 175

Convert rough user ideas/proposals/chat transcripts into structured TODO DSL steps, with per-step knowledge/codebase references for low-bias orchestration.

CurryTang By CurryTang schedule Updated 3/3/2026

name: todo-dsl-generator description: Convert rough user ideas/proposals/chat transcripts into structured TODO DSL steps, with per-step knowledge/codebase references for low-bias orchestration.

TODO DSL Generator

Use this skill when a user provides:

  • rough idea bullets,
  • long proposal text,
  • chat logs from Claude/GPT, and wants actionable structured steps.

Input

  • user instruction or proposal text
  • project metadata (name, projectPath, knowledgeGroupIds)

Output Contract

Return todoDsl with ordered steps:

  • step_id
  • title
  • kind
  • objective
  • assumptions[]
  • acceptance[]
  • commands[]
  • checks[]
  • depends_on[]
  • references.knowledge[]
  • references.codebase[]

Also return flattened todoCandidates[] for TODO cards.

Reference Policy

For each step, bind references from:

  1. knowledge base assets/documents
  2. project codebase files

If no strong match exists, keep references empty rather than fabricating.

Quality Gates

  • avoid generic placeholders
  • one step should be runnable/reviewable in one session
  • dependency order must be explicit
  • acceptance/check definitions should be concrete
Install via CLI
npx skills add https://github.com/CurryTang/Amadeus --skill todo-dsl-generator
Repository Details
star Stars 175
call_split Forks 17
navigation Branch main
article Path SKILL.md
More from Creator