agentleFS
Sign inStart free
OKHP3OKHP3/skillz/.agents/skills/okhp3-process-capture/SKILL.md

okhp3-process-capture

Skill3 starsChanged 2 months agoMITView on GitHub

Capture a recurring task as either a backlog entry or a new skill skeleton. Use when the user says "I keep doing X", "this is the third time", "make this a skill", "let's capture this", or at the natural end of a work session when a repeatable pattern just occurred. Extracts the pattern from conversation context (tools used, sequence of steps, input/output formats, corrections made) and produces either a BACKLOG.md line or a SKILL.md skeleton following this repo's conventions — not full skill authoring (that's skill-creator's job once a skeleton exists).

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Update the index: Add the new skill's trigger row to root AGENTS.md, even though the skeleton isn't fully authored yet — mark it in the …

    line 34

≈ “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

okhp3-process-capture

OverKill Hill P³ · overkillhill.com · github.com/OKHP3

The intake mechanism. Turns "I just did this for the third time" into a structured artifact, not a lost observation.

Process

  1. Extract the pattern. From the current conversation (or a description the user gives), identify: what was the task, what tools/skills were used, what was the input, what was the output, were there corrections or judgment calls along the way that would need to be encoded as instructions.
  2. Assess readiness. Has this pattern stabilized enough to write down, or is this still "developing taste" (per the original delegation-contract framing — genuinely novel judgment calls each time means it's not a skill yet)?
    • Not ready -> write a docs/BACKLOG.md entry (date, task, times observed, candidate family, status: captured). Stop here.
    • Ready -> continue to step 3.
  3. Determine target family. Does this fit an existing family (mermaid/, social-posting/, or this repo's own process-capture/), or does it need a new top-level family? If new, note this explicitly — adding a new family directory is a bigger decision than adding a skill within one, and the user should confirm.
  4. Produce the skeleton. Frontmatter (name, "pushy" description per skill-creator's triggering guidance) + section-header outline matching the body's intended structure, with 1-2 sentence descriptions per section and pointers to reference files that will need authoring. Match the structure of existing skills in this repo (core+domain pattern for multi-skill families, single-skill pattern for self-contained ones).
  5. Update the index. Add the new skill's trigger row to root AGENTS.md, even though the skeleton isn't fully authored yet — mark it in the "Planned" section if the body is still mostly TOC, move to "Active" once authored.

Handoff

A skeleton from this skill is a starting point for skill-creator (the Anthropic example skill, if available) for full authoring, testing, and description-optimization — this skill does intake and structure, not iteration and evals.

About

Built by Jamie Hill · OverKill Hill P³ Published at github.com/OKHP3 Part of the OKHP3/skillz Agent Skill library. MIT License -- free to use, fork, and adapt. A nod to the source is appreciated.

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