agentsclimarketplace

Skill authoring

Skill martonpaulo/skills/skill-authoring

My personal Agent Skills for coding agents: engineering workflows plus a few personal-use tools. Personalized forks, with every upstream author credited.

Install
npx -y skills add martonpaulo/skills --skill skill-authoring

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 14 days oldThe repository was created 14 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Create, review, or simplify Agent Skills with precise triggers, focused workflows, explicit safety boundaries, useful completion criteria, valid frontmatter, and minimal unnecessary context.

SKILL.md

2.4 KB, as published. Nobody here has run it

Skill Authoring

Use this user-invoked skill to create a new Agent Skill or improve an existing one. Optimize for predictable behavior, narrow responsibility, and low context cost.

Workflow

  1. Inspect the existing skill, supporting files, nearby skills, and repository guidance. Preserve useful behavior and identify stale references before editing.
  2. Define the skill contract:
    • Objective: the single outcome the skill owns.
    • Prerequisites: evidence, files, tools, or state required to begin.
    • Invocation: who should trigger it and under what narrow conditions.
    • Workflow: ordered actions that materially change agent behavior.
    • Safety boundaries: destructive, publishing, Git, data, and scope limits specific to the skill.
    • Completion criteria: observable conditions for a finished run.
    • Validation: structural and behavioral checks for the skill itself.
  3. Choose invocation policy:
    • Model-invoked: omit disable-model-invocation; write a narrow description containing positive trigger conditions and explicit non-triggers when accidental activation is plausible.
    • User-invoked: set disable-model-invocation: true; write a concise human-facing description.
  4. Put the common workflow in SKILL.md. Move branch-specific reference material to clearly named supporting files only when that reduces the context needed for ordinary runs.
  5. Remove no-op instructions, promotional prose, stale assumptions, and generic rules already owned by repository guidance. Keep each meaning in one place.
  6. Use cross-skill references only for real dependencies, and verify that every referenced skill is installed. Avoid tool-specific instructions unless the skill genuinely requires that tool.
  7. Add examples only when they materially disambiguate behavior. Keep them short and adaptable.
  8. Run the checks in VALIDATION.md and review the final diff.

Completion

The skill is complete when its name and directory match, invocation is correct, references resolve, steps and boundaries are operational, completion is checkable, and unnecessary context has been removed.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.