agentleFS
sethatwoodsethatwood/belay/skills/learn/SKILL.md

learn

Skill0 starsChanged 51 days agoMITView on GitHub

Start learning. In an empty folder, Belay asks what you want to learn and proposes three real projects. In an existing codebase, it maps the repo, finds the skills you haven't earned, and proposes real changes that teach them.

195 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use add-anydoc, add-clidash and 10 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.
  • TomGranotSkill

    add-anydoc

    Add local office-document-to-Markdown conversion to NanoClaw agent containers with the pinned Firecrawl AnyDoc CLI. Use when agents need to read attached Word, PowerPoint, Excel, OpenDocument, RTF, EPUB, CSV, or text-based PDF files without uploading them to a hosted parser.

    Also in 5 projects that use this

  • TomGranotSkill

    add-clidash

    Add clidash — a zero-dependency, read-only web dashboard that derives its tabs and tables at runtime from any CLI that lists resources as JSON. Ships pre-wired for NanoClaw's ncl CLI (agent groups, sessions, channels, users, roles), plus message-activity charts, a log tail, and a read-only file viewer for group skills/CLAUDE.md/profiles.

    Also in 5 projects that use this

  • seek888Skill

    open-gstack-browser

    Launch GStack Browser — AI-controlled Chromium with the sidebar extension baked in. Opens a visible browser window where you can watch every action in real time. The sidebar shows a live activity feed and chat. Anti-bot stealth built in. Use when asked to "open gstack browser", "launch browser", "connect chrome", "open chrome", "real browser", "launch chrome", "side panel", or "control my browser". Voice triggers (speech-to-text aliases): "show me the browser".

    Also in 11 projects that use this

  • Saqlain-811Skill

    ios-sync

    Regenerate the iOS debug bridge against the latest upstream gstack templates. (gstack)

    Also in 6 projects that use this

  • MicrosoftSkill

    add-codex

    Use Codex (CLI + AppServer) as the full agent provider — planning, tool orchestration, native compaction, MCP tools, session resume — in place of the Claude Agent SDK. ChatGPT subscription or OPENAI_API_KEY. Per-group via agent_provider. Distinct from using OpenAI as an MCP tool (where Claude remains the planner).

    Also in 6 projects that use this

  • MicrosoftSkill

    add-deltachat

    Add DeltaChat channel integration via @deltachat/stdio-rpc-server. Native adapter — no Chat SDK bridge. Email-based messaging with end-to-end encryption.

    Also in 6 projects that use this

  • +5 more from TomGranot
  • +1 more from microsoft

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 170 edited versions we read keep these.

  • 6

    Distinctive

    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. Distinctive:

    Look at the working directory.

    line 11
  2. Very specific:

    If the folder already has .belay/map.json, skip belay_init in either case: call belay_map and work from what is unearned.

    line 13
  3. Distinctive:

    Ask what they want to learn: a language, a stack, or a thing they want to be able to build.

    line 17
  4. Distinctive:

    Ask how much time they have for it, and whether there is a date they need it by.

    line 18
  5. Distinctive:

    Ask two or three questions about their life, one at a time, to find a program they would actually use.

    line 19
  6. Very specific:

    When they pick, call belay_init with map set to typescript or python, whichever fits what they said.

    line 21
  7. Distinctive:

    Ask what they want to learn here.

    line 29
  8. Distinctive:

    Read the repo: the package manifest or requirements, the test setup, the routes, the migrations, the jobs.

    line 30
  9. Very specific:

    When they pick, call belay_init with the language's map and the precedents you found, then belay_begin_step on the change's skill, and …

    line 33

“Common” is measured: kept by at least 6 in 10 of the 170 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

The person wants to learn something. Find out what, set the repo up, and open the first step. Everything in the Belay contract applies from the moment the map exists.

First, which kind of folder this is

Look at the working directory. If it holds no source files, this is a fresh start. If it holds a codebase, this is learning inside real work. Do not ask which; look.

If the folder already has .belay/map.json, skip belay_init in either case: call belay_map and work from what is unearned.

A fresh start

  1. Ask what they want to learn: a language, a stack, or a thing they want to be able to build. One question.
  2. Ask how much time they have for it, and whether there is a date they need it by. One question.
  3. Ask two or three questions about their life, one at a time, to find a program they would actually use. Something they track by hand, a chore they repeat, a thing they always forget.
  4. Propose three projects, all real, each sized to the time they have, each labeled with the skills it teaches in the order it meets them. Number them and ask them to pick one. No more than three sentences per project.
  5. When they pick, call belay_init with map set to typescript or python, whichever fits what they said. If it reports gitInit, say in one line that the folder is now a git repo, because Belay reads their work from git.
  6. If what they want to learn needs skills the map does not name, add them to .belay/map.json now, before any step is open, by editing the file: same shape as the others, id in kebab-case, teaches in the one or two sentences a senior would say, witness from test, build, and types, requires naming only skills above it, precedents empty. Add only skills a test, a build, or a type check can witness, and only as many as the project will reach. Name the skills you added in one line.
  7. Say the map's first skill is writing a test, every skill starts unearned, and the first step begins now. Call belay_begin_step on write-a-test with a goal that is the first failing test of their project, and follow the contract: the goal in two sentences, then ask what they would do first.

Belay supports TypeScript, JavaScript, and Python in version one. If they name another language, say so plainly and offer the nearest of those.

Inside a codebase

  1. Ask what they want to learn here. One question. If they do not know, say you will look for the skills this codebase uses that they have not earned yet.
  2. Read the repo: the package manifest or requirements, the test setup, the routes, the migrations, the jobs. Match what you find to the skills on the starter map for the repo's language.
  3. For each skill the repo uses, find one or two files that show the pattern done well. Those are the precedents.
  4. Propose up to three real changes that teach skills they have not earned: a test that is missing, a pattern the codebase uses that they have not done yet, a gap someone senior would have fixed. Each one points at the file where it was done before. If there is no real place in this repo to learn a skill, say so; never invent work.
  5. When they pick, call belay_init with the language's map and the precedents you found, then belay_begin_step on the change's skill, and follow the contract.

Always

Short turns. One question at a time. Do not explain how any project would be built; the building is theirs.

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.