agentleFS
Sign inStart free
expoexpo/exact/.claude/skills/llp-create/SKILL.md

llp-create

Skill4.3k starsChanged 3 days agoMITView on GitHub

Create a new LLP document with the next available number (top-level NNNN or a dotted sub-LLP under a parent), the NNNN-slug.type.md filename convention, and a scaffolded metadata header, linked into llp/current/ — or extend an existing LLP when one already covers the topic.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 1

    1. expo/expo-device-hub★ 285100% same

    Edited · 0

    None found.

    Wrote their own version · 1

    1. hyparam/hypaware★ 16same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository9 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 2 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

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

    MUST check for a covering document first: Grep the Systems: and Related: headers across llp/ and propose *extending* an existing LLP over …

    line 22
  2. Very specific:

    MUST name the file NNNN(.NNN)*-slug.type.md with the lowercased type matching the header's Type:, put the full dotted number in the title, …

    line 24
  3. Very specific:

    MUST link the new document into llp/current/ with a relative symlink (mkdir -p llp/current && ln -s ../<file> llp/current/<file>.

    line 25
  4. Very specific:

    MUST propose a sub-LLP, not a new top-level number, when the document refines, specifies, or plans an existing LLP (spec for an RFC → …

    line 26
  5. Distinctive:

    MUST NOT overwrite an existing file — if the name collides, stop and ask.

    line 27
  6. Distinctive:

    MUST NOT silently pick a document type when the choice is ambiguous — propose one and confirm.

    line 28
  7. Very specific:

    No corpus to add the document to → llp-adopt.

    line 40

≈ “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

llp-create

<!-- @ref https://github.com/ccheever/llp/blob/v0.5.2/llp/0000-linked-literate-programming.explainer.md#1-llp-documents — numbering, metadata, types, statuses --><!-- @ref https://github.com/ccheever/llp/blob/v0.5.2/llp/0008-distributed-agent-skills.rfc.md#the-skills--core-five-plus-flags — skill contract -->

Author one LLP in a corpus that already exists. Handles numbering, filename, and the metadata scaffold so the author can focus on content.

Invoke as /llp-create <title>, /llp-create --under NNNN <title> for a sub-LLP, or bare and the skill asks.

Trigger

"Capture this decision" / "write an LLP for this" — during or after a design conversation.

Invariants

  • MUST check for a covering document first: Grep the **Systems:** and **Related:** headers across llp/ and propose extending an existing LLP over creating a new one when the topic is already covered.
  • MUST derive the next number as max(ever allocated) + 1 — top-level NNNN, or for a sub-LLP of P: P. + max(ever allocated children of P) + 1 as three digits (000 if none) — where "ever allocated" scans all of llp/ recursively (excluding current/, foundation/, reviews/) and history for deleted documents (git log --all --diff-filter=D --name-only -- llp). Numbers are never reused, never invented, and a parent must already exist.
  • MUST name the file NNNN(.NNN)*-slug.type.md with the lowercased type matching the header's **Type:**, put the full dotted number in the title, and start every new LLP as Draft.
  • MUST link the new document into llp/current/ with a relative symlink (mkdir -p llp/current && ln -s ../<file> llp/current/<file>; on a checkout without symlink support, a regular file containing ../<file> is equivalent) — new work is in the working set by definition (LLP 0000 Current and foundation).
  • MUST propose a sub-LLP, not a new top-level number, when the document refines, specifies, or plans an existing LLP (spec for an RFC → NNNN.000-….spec.md). Never re-parent by renaming.
  • MUST NOT overwrite an existing file — if the name collides, stop and ask.
  • MUST NOT silently pick a document type when the choice is ambiguous — propose one and confirm.

Workflow

Recipe (advisory) — next top-level number: { find llp -name '[0-9]*.md' -not -path 'llp/reviews/*' -not -path 'llp/current/*' -not -path 'llp/foundation/*'; git log --all --diff-filter=D --name-only --format= -- llp; } | grep -oE '(^|/)[0-9]{4}' | grep -oE '[0-9]{4}' | sort -n | tail -1 (+1); next child of 0011: same pipeline with 0011\.[0-9]{3} (+1, or 000). Slug: lowercase, non-alphanumerics → -, collapse repeats, keep it under ~6 words. Type cues: proposal → rfc; requirements → spec; settled choice → decision; steps → plan; teaching → explainer; always/never → principles; how-to → guide; bug → issue; findings → research. Scaffold sections: Summary · Motivation · Design · Open questions (Decisions use Context/Options/Decision/Consequences; Research uses Findings/Confidence).

Artifact

A new Draft LLP with a complete metadata header (Type, Status, Systems, Author, Date; Related: where relevant) — plus an offer to draft the body from the current conversation. Confirm the assigned number and path back to the user.

Hand-offs

  • No corpus to add the document to → llp-adopt.
  • The document is foundational or contentious → llp-review, scaled to stakes (LLP 0005).

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