agentleFS
Sign inStart free
KevoyuanKevoyuan/AgentSCAD/skills/developer-workflow/SKILL.md

developer-workflow

Skill26 starsChanged 5 months agoMITView on GitHub

Work safely on the AgentSCAD codebase, especially skill/docs changes around the CAD pipeline. Use when editing AgentSCAD skills, docs, resolver guidance, or orchestration-adjacent notes while preserving runtime contracts and avoiding unrelated code changes.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 10 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this workflow for codebase work that touches AgentSCAD skills or documentation.

    line 13
  2. Distinctive:

    Read the relevant skill or doc before editing.

    line 18
  3. Distinctive:

    If runtime behavior matters, read the route or helper that owns the contract, but do not edit it unless explicitly requested.

    line 19
  4. Distinctive:

    Keep deterministic behavior in TypeScript/Python tools, not prompts.

    line 34
  5. Very specific:

    Keep SKILL.md concise and task-oriented.

    line 38
  6. Distinctive:

    Put trigger guidance in the description.

    line 40
  7. Distinctive:

    Prefer imperative workflow instructions over broad background.

    line 41
  8. Distinctive:

    Do not duplicate large source snippets when a short contract summary is enough.

    line 42
  9. Distinctive:

    Do not revert other people's edits.

    line 46
  10. Distinctive:

    Keep changes scoped to requested files.

    line 47
  11. Very specific:

    Avoid touching src/, package files, Prisma files, or existing CAD skills unless explicitly authorized.

    line 48
  12. Distinctive:

    Use deterministic tests and wrapper tools instead of guessing about runtime behavior.

    line 49
  13. Distinctive:

    After edits, verify with git diff --stat and inspect the changed files.

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AgentSCAD Developer Workflow

Use this workflow for codebase work that touches AgentSCAD skills or documentation.

First Read

  1. Inspect git status --short.
  2. Read the relevant skill or doc before editing.
  3. If runtime behavior matters, read the route or helper that owns the contract, but do not edit it unless explicitly requested.

Protected Contracts

Preserve these exactly unless the user explicitly requests a coordinated runtime change:

  • SSE data frames: data: ${JSON.stringify(payload)}\n\n.
  • State strings: NEW, SCAD_GENERATED, RENDERED, VALIDATED, DELIVERED, DEBUGGING, REPAIRING, VALIDATION_FAILED, GEOMETRY_FAILED, RENDER_FAILED, HUMAN_REVIEW, CANCELLED.
  • Step strings: starting, generating_llm, generating_mock, scad_generated, scad_applied, rendering, render_failed, rendered, validating, validation_failed, validated, delivering, delivered.
  • Artifact paths: /artifacts/{jobId}/model.scad, /artifacts/{jobId}/model.stl, /artifacts/{jobId}/preview.png, optional /artifacts/{jobId}/report.
  • Generation JSON: summary, parameters, scad_source.
  • Parameter fields: key, label, kind, unit, value, min, max, step, source, editable, description, group.
  • Validation fields: rule_id, rule_name, level, passed, is_critical, message.
  • Provider fallback: MiMo when configured, then ZAI SDK. Product generation fails visibly when no model succeeds; legacy templates are available only through the explicit demo-only switch.
  • Rendering and validation: OpenSCAD CLI renders artifacts; Python/trimesh validates STL with mock fallback.
  • Keep deterministic behavior in TypeScript/Python tools, not prompts.

Skill Writing

  • Keep SKILL.md concise and task-oriented.
  • Include YAML frontmatter with name and description.
  • Put trigger guidance in the description.
  • Prefer imperative workflow instructions over broad background.
  • Do not duplicate large source snippets when a short contract summary is enough.

Editing Rules

  • Do not revert other people's edits.
  • Keep changes scoped to requested files.
  • Avoid touching src/, package files, Prisma files, or existing CAD skills unless explicitly authorized.
  • Use deterministic tests and wrapper tools instead of guessing about runtime behavior.
  • After edits, verify with git diff --stat and inspect the changed files.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free