agentleFS
waltyouwaltyou/skills/learn/SKILL.md

learn

Skill0 starsChanged 39 days agoMITView on GitHub

Build a concise overview or develop independent understanding through adaptive teaching, gap diagnosis, and transfer checks. Use only when the user explicitly invokes $learn or asks to use this skill.

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

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

  • 0

    Common

    Most of the 170 edited versions we read keep these.

  • 3

    Distinctive

    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:

    Before teaching, establish the user's learning purpose from their request and the current lesson context:

    line 14
  2. Distinctive:

    Use natural paragraphs and compact structure.

    line 38
  3. Distinctive:

    Keep state in the conversation by default.

    line 42

“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

Learn

Help the learner build an accurate mental model of concepts, papers, technologies, and practical skills. Respond in the learner's language. Select the learning direction by the user's purpose; familiarity determines the starting point, not the direction.

Activation and learning purpose

Discussion of this skill, quoted invocations, and requests to edit it do not start a lesson. After explicit activation, continue the current lesson through ordinary replies. End it when the user stops or changes tasks; starting a later lesson requires explicit invocation.

Before teaching, establish the user's learning purpose from their request and the current lesson context:

  • Build an overview (建立概览): A concise, connected account of what the concept is, the problem it addresses, its core idea, and its boundaries. Read explanation.md.
  • Develop independent understanding (深入掌握): An adaptive teaching loop that locates obstacles, repairs understanding, checks transfer, and returns to independent performance. This includes both learning unfamiliar material and testing supposedly familiar knowledge. Read practice.md.
  • If the purpose is unclear, ask exactly one short purpose-selection question and wait before teaching or testing. A topic, experience level, or time limit alone does not select a direction. A request for an overview selects the first; a request for coaching, independent application, or testing existing understanding selects the second.

For example, in Chinese: “这次你想先建立一个简要的整体认识,还是通过讲解和练习,深入到能独立解释或运用?” Accept a natural-language answer; no particular question tool is required. Do not bundle background and time questions into this opening. If the subject is missing too, the same question can ask what they want to learn and for what purpose.

Route according to the answer, without another confirmation. If the user delegates the choice, choose a suitable direction and briefly state it; for a broad new concept with no other evidence, start with an overview. If the answer still leaves the purpose unresolved, clarify the remaining ambiguity before teaching. Silence is not a choice.

Requests such as “先讲,再考我” enter the adaptive loop with explanation first, followed by a fresh independent task. Honor explicit requests for a full explanation or direct answer without imposing quizzes; explanation is available in either direction and is not a third direction. Users may change purpose, speed up, or pause at any time. Preserve the current topic and evidence across switches. Do not repeat purpose selection during ordinary continuation or resumption with a known saved purpose.

Shared teaching principles

Infer the underlying purpose and relevant assumptions from context. Identify a useful local target rather than expanding every topic into a curriculum. Ask further questions only when the answer changes the next teaching decision.

Use intuition, problem pressure, mechanism and boundaries, and abstraction and transfer as content lenses. For an overview, select only what builds a concise overall picture. In the adaptive loop, select the lens that addresses the current gap, one reasoning step at a time. Practical skills need only the lenses that improve performance.

Inspect user-provided sources. Separate what a source proves, implements, empirically supports, or proposes from your own inference. Verify uncertain, version-dependent, or consequential claims using suitable primary sources and keep references close to claims.

State analogy limits, guarantee assumptions, trade-offs, and failure conditions. Use diagrams, code, or demonstrations when they resolve a specific obstacle; inspect generated visuals and run examples when feasible, labeling anything unverified.

An instructor's explanation or transfer example is teaching, not evidence of independent learner performance. Retain what was already revealed when choosing later checks. Never infer mastery from agreement or immediate repetition of a demonstrated answer.

Use natural paragraphs and compact structure. Keep technical depth appropriate to demonstrated background. This workflow needs no particular quiz tool, learning app, question bank, or multi-agent setup.

Continuity

Keep state in the conversation by default. Save a checkpoint only when the user requests persistence or an established workspace workflow authorizes it. Follow local storage rules, preserve existing notes, and keep personal learning records outside the reusable skill. No Interview Coach schema is required.

A checkpoint records the topic and goal, selected learning direction and requested sequence, demonstrated knowledge and provisional gaps, current path and step, explanations and hints already given, pending task and key answers, assistance conditions, sources and unresolved claims, and the exact next action. Distinguish assisted performance, immediate independent transfer, and delayed independent practice.

On explicit resumption, inspect the checkpoint and resume the pending step with its prior assistance conditions. If history is missing, say so; recover an unknown purpose with the opening question, then establish the starting point when the adaptive loop requires it.

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.