agentleFS
ai-erp-collabai-erp-collab/deterministic-context-routing-plugin/skills/curate/SKILL.md

curate

Skill0 starsChanged 4 days agoMITView on GitHub

Use when the user asks, in a named module, to write or update a knowledge-base page (concept, table, contract, usage flow, or pattern) from a raw artifact, a schema or table export, or code, for example "write a contract page for X", or to mark an artifact processed. Not for answering questions from existing pages or registering modules.

10 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use learn, triage and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • genomewalkerSkill

    learn

    Learn from web sources, generate SKILL.md, and store key patterns in soul memory. Trigger keywords: learn, skill-gen, web-learn, documentation, tutorial.

    Also in 3 projects that use this

  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 1 is distinctive, and 7 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 7

    Very specific

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

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

    If tools/kb/kb.py is missing but AGENTS.md and modules.md exist, the workspace predates the integrity tool: tell the user to run the init …

    line 8
  2. Very specific:

    If the module has no concept.md yet, write it first.

    line 28
  3. Very specific:

    Look the subject up first: python tools/kb/kb.py resolve "<name>".

    line 29
  4. Very specific:

    If a page for it exists, update that page instead of creating a second one.

    line 30
  5. Very specific:

    Write the page under knowledge-base/modules/<module>/.

    line 42
  6. Distinctive:

    Never edit source code while curating.

    line 54
  7. Very specific:

    Add or update one line for the page in the module INDEX.md: path and the question the page answers.

    line 61
  8. Very specific:

    Do not touch the artifact or source-artifacts/INDEX.md.

    line 63

“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “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

Curate

If tools/kb/kb.py is missing but AGENTS.md and modules.md exist, the workspace predates the integrity tool: tell the user to run the init skill, which upgrades it, and stop. Run tool commands from the workspace root; use python3 or py -3 where python is missing or older than 3.9.

Before writing

  1. Resolve the module in modules.md. Read its AGENTS.md, its knowledge-base INDEX.md, and its inventory.md.
  2. Choose the page type and template from templates/wiki/ in the workspace:
    • table: a schema, export, or data structure → tables/<name>.md;
    • contract: a function, class, or API that others call → contracts/<name>.md;
    • usage flow: a user or system flow across steps → usage/<name>.md;
    • pattern: one technical shape repeated in two or more places → patterns/<name>.md. There is no pattern template: use the usage flow template and list every place the shape occurs;
    • concept: the module's purpose, boundaries, and terms → concept.md. If the module has no concept.md yet, write it first.
  3. Look the subject up first: python tools/kb/kb.py resolve "<name>". If a page for it exists, update that page instead of creating a second one. Facts that serve several modules go to knowledge-base/shared/.
  4. Large inputs: read a source file over about 50 KB through python tools/kb/kb.py outline <file> and then only the span you need. A table export goes through python tools/kb/kb.py table <export> --module <module> (the kb skill's table procedure covers a layout the tool does not know yet). Evidence that is not in source-artifacts/ yet is taken in first with python tools/kb/kb.py ingest <path> --module <module>.

Writing

  1. Write the page under knowledge-base/modules/<module>/. Every non-obvious claim ends with a confidence label:
    • confirmed: seen in code, data, or a document you read;
    • inferred: concluded from evidence, not stated anywhere;
    • runtime-unverified: depends on behavior you could not run;
    • blocked-by-access: the evidence exists but you cannot reach it. Set the page-level confidence: to the weakest label on the page.
  2. Required sections: a table page has ## Keys, ## Fields, and ## Relations; a class or function contract has ## Purpose and ## Signature.
  3. The page's source: line names the artifact copy under source-artifacts/ it was written from.
  4. Never edit source code while curating.

After writing

  1. In inventory.md, end each line the page now covers with -> <page path>, or with excluded: <reason> when it deliberately stays out.
  2. Add or update one line for the page in the module INDEX.md: path and the question the page answers.
  3. Do not touch the artifact or source-artifacts/INDEX.md. Once the page names the copy in source:, the write hook (or the next python tools/kb/kb.py fix --safe) records the artifact as processed in the registry. Files stay where they are.

Finish with the close-session skill.

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.