agentleFS
expoexpo/exact/.claude/skills/llp-adopt/SKILL.md

llp-adopt

Skill4.3k starsChanged 3 days agoMITView on GitHub

Bring LLP to a repository and keep its installed skills current — the single entry point for "set up LLP here" and "update the LLP skills". Scaffold mode (fresh structure), retrofit mode (survey the codebase, draft a provenance-tagged root LLP, propose an adoption plan), install step (copy skills from a tagged release with a receipt), and update mode (verified, diff-approved skill updates).

2 other projects use a version of this, 1 of them edited.

Trending

More talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 0 mentionsNov 2026: 19 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the expo/exact repository that our weekly crawl found. There is no adoption data for this file yet.

Most recently taken up by

Who's talking about expo/exact

The biggest 5 of 19 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 19

Talked about on GitHub, Bluesky and Hacker News (19 times).

On GitHub · 11

6 more on GitHub

On Bluesky · 6

1 more on Bluesky

On Hacker News · 2

  • Other versions2 projects use a version of this.

    Word for word · 1

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

    Edited · 1

    1. peterpme/skills★ 2895% same

    Wrote their own version · 0

    None found.

    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

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    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:

    MUST extend an existing llp/ tree if one exists — never start a second setup layer.

    line 21
  2. Distinctive:

    MUST write only new files and clearly-marked managed blocks.

    line 22
  3. Very specific:

    MUST keep every generated document Draft and provenance-tagged: claims are [observed] (say where), [confirmed] (named human + date), or …

    line 23
  4. Very specific:

    MUST propose @ref annotations and apply them only after approval — no mass annotation, and none on volatile early code.

    line 24
  5. Distinctive:

    MUST NOT create a batch of speculative LLPs in either mode.

    line 25
  6. Distinctive:

    When a tool expects CLAUDE.md, symlink it to AGENTS.md when safe rather than maintaining divergent copies.

    line 26
  7. Distinctive:

    MUST confirm before writing user/global skill directories (they affect every repo for that user).

    line 31
  8. Distinctive:

    Update MUST present the full changeset for the installed profile *as defined at the target tag* as one diff, revalidate the approved state …

    line 33
  9. Distinctive:

    If upstream is unreachable, a tag moved or vanished, or content can't be verified against the resolved commit: report exactly that and …

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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-adopt

<!-- @ref https://github.com/ccheever/llp/blob/v0.5.2/llp/0001-adopting-llp.guide.md — the workflow this automates --><!-- @ref https://github.com/ccheever/llp/blob/v0.5.2/llp/0010.000-skill-installation-and-updates.spec.md — install/update design, receipts, trust model --><!-- @ref https://github.com/ccheever/llp/blob/v0.5.2/llp/0000-linked-literate-programming.explainer.md#provenance-for-generated-rationale — tagging rules for generated docs -->

One entry point for "add LLP to this project" and "update the LLP skills". Invoke as /llp-adopt (auto-detect scaffold/retrofit), --mode scaffold / --mode retrofit, or /llp-adopt update.

Trigger

"Set up / bring LLP to this repo," for any repository state; "update the LLP skills," in any repo with installed skills.

Invariants — adoption

  • MUST extend an existing llp/ tree if one exists — never start a second setup layer.
  • MUST write only new files and clearly-marked managed blocks; existing user prose is never overwritten without showing a diff and asking. Reruns MUST be idempotent (no duplicate files or blocks).
  • MUST keep every generated document Draft and provenance-tagged: claims are [observed] (say where), [confirmed] (named human + date), or [inferred] — and [inferred] claims never survive into a document proposed for promotion.
  • MUST propose @ref annotations and apply them only after approval — no mass annotation, and none on volatile early code.
  • MUST NOT create a batch of speculative LLPs in either mode; establish orientation and the loop, then hand off.
  • When a tool expects CLAUDE.md, symlink it to AGENTS.md when safe rather than maintaining divergent copies.

Invariants — install and update (LLP 0010.000)

  • MUST install only from a tagged release (the tag's tree is the release), as copies never symlinks, and write a receipt beside the installation: origin, tag, resolved commit, profile flags (core; review-plus; delivery, which also records the repo's delivery policy in repo-local config), and a per-file content hash. The receipt — not source: frontmatter — is the trust root for updates.
  • MUST confirm before writing user/global skill directories (they affect every repo for that user); project-local may proceed more freely.
  • Update MUST three-way-check each file against the receipt hash. Any mismatch is a local fork: abort the coupled update, never clobber, and offer discard (restore managed copy) or detach (file leaves the managed set permanently; receipt records it; later updates ignore it and note upstream changes without applying).
  • Update MUST present the full changeset for the installed profile as defined at the target tag as one diff, revalidate the approved state immediately before mutation, and be recoverable to either endpoint (the receipt write is the commit point). Runs are single-flight per installation.
  • If upstream is unreachable, a tag moved or vanished, or content can't be verified against the resolved commit: report exactly that and write nothing. Cannot-verify never degrades to guess.

Workflow

Recipe (advisory) — Adoption: inspect first (existing llp/? agent files? how much code?): little to mine → scaffold; substantial history → retrofit. Both share the spine: llp/ with its current/ and foundation/ overlay directories, root LLP 0000 (Explainer, **Role:** Root, Draft — linked into current/; it enters foundation/ when the author makes it Active), the AGENTS.md managed block. Retrofit adds: survey, draft root LLP, propose 2–5 subsystem LLPs, layered annotation plan (LLP 0001). Install: offer after the managed blocks; default profile core, flags opt-in. Update: fetch the tag's tree (shallow clone or archive; git required), diff, approve, apply as a staged set with prior bytes retained until the receipt lands; rollback is git revert where the installation is committed, the backups otherwise. Releasing (this repo): bump source: pins and pinned URLs → commit → tag → push tag — a skill-file change is not installable until tagged.

Artifact

Scaffold: llp/ with current/ (holding the link to the new root) and an empty foundation/, an LLP 0000 skeleton, the managed blocks. Retrofit: those plus a survey summary, Draft provenance-tagged document(s), and a proposed adoption plan. Install/update: pinned skill copies, the receipt, and (with the delivery flag) the repo policy file.

Hand-offs

  • Corpus exists → daily work starts with llp-orient; author follow-on docs with llp-create.
  • Ratify generated drafts with llp-review where stakes warrant; keep healthy with llp-maintain — whose audit notes stale pins and hands back to /llp-adopt update.

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.