About ub-authoring
A reusable authoring skill for creating portable guidance for skills and related agent-behavior surfaces. It defines routing, boundaries, references, and validation expectations while keeping domain-specific material in its owning skill.
Key features
- 01Portable skill authoring conventions
- 02Routing and non-use boundaries
- 03On-demand reference loading
- 04Shared choice-question guidance
- 05Output and completion checklists
Use cases
- 01Shaping a reusable skill structure
- 02Improving skill descriptions and routing
- 03Setting boundaries to prevent misrouting
- 04Reviewing portable skill content
Install
Copy the command and run it in a terminal.
npx skills add https:/ / github.com/ robert-hoffmann/ uncle-bob --skill ub-authoringAfter installing, type/ub-authoringin Claude Code to confirm it loaded.
Bundled filesreferences/authoring-conventions.md
Details
From the project READMEUB Authoring
Mission
Use this skill to shape portable, installable authoring guidance for skills and related agent-behavior surfaces.
This skill owns the reusable authoring conventions that should travel with the skill catalog itself instead of living in repo-only docs or in one customization-specific skill.
When Not To Use
- Do not use this skill for end-to-end customization generation; defer that to
ub-customizations. - Do not use this skill for ordinary code implementation or framework work where no shared authoring contract is being changed.
- Do not use this skill as a substitute for repo-local maintenance docs or packaging metadata that belong to a development repository rather than the distributable skill payload.
Embedded Contract
These rules are the base contract of this skill and must not depend on a secondary document to be applied correctly.
- Treat descriptions as routing metadata, not marketing copy.
- Add explicit non-use boundaries when wrong-skill activation is realistically costly or common.
- Prefer concrete examples and acceptance boundaries over vague admonitions.
- Keep the main skill surface lean and move deeper detail into targeted references.
- Keep shared authoring guidance portable across installed skill payloads and avoid repo-local live truth in the reusable contract.
- Keep reusable choice-question UX contracts in shared authoring guidance instead of re-encoding the same multiple-choice rules in each skill.
Load References On Demand
- Read
references/authoring-conventions.mdwhen shaping shared skill structure, cross-skill wording, routing quality, reusable choice-question UX, or other reusable authoring patterns.
Core Workflow
- Decide whether the task is changing a reusable authoring convention or a single skill's local content.
- Keep portable shared guidance in this skill and domain-specific guidance in the owning skill.
- Normalize descriptions, examples, non-use boundaries, and shared choice-question behavior only where the changed surfaces actually benefit from them.
- Keep reusable references installable and avoid dead links into repo-only docs.
Output Requirements
When producing non-trivial authoring updates, include:
routing_noteboundary_notereference_notevalidation_note
Completion Checklist
- Shared guidance remains portable inside
.agents/skills/. - Description wording reads like routing logic, not inventory prose.
- Non-use boundaries are explicit where misrouting is realistic.
- Examples are concrete where abstract guidance would be easy to misapply.
- Shared references avoid repo-only assumptions unless clearly labeled as such.