agentleFS
contract-herocontract-hero/agentic-community-college/skills/course-creator/SKILL.md

course-creator

Skill0 starsChanged 2 months agoMITView on GitHub

Scaffold a new ACC content plugin (a "course") from scratch. Emits a Claude Code plugin manifest that declares accContent.lessons + accContent.probes, plus a start command, README, CLAUDE.md, .gitignore, and an empty lessons/ directory. Walks the user through declaring the course's prerequisite probes with the four supported kinds (filesystem-exists, http-get, shell-exit-zero, claude-plugin-enabled). Use when the user asks to "create a new course", "scaffold a course plugin", "start a new ACC course", "add a domain to ACC", or wants a fresh content plugin before authoring any lesson.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. aiunlocked1412/claude-skill-unlock★ 18same name
    2. davekindl/skills★ 2same name
    3. GarphenGate/moltline-oss★ 0same 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 repository4 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, 3 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

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

    Ask the user where the new course should live.

    line 14
  2. Distinctive:

    Author: default to git config user.name, ask to confirm.

    line 24
  3. Distinctive:

    Ask the user: "What prerequisites does any lesson in this course need?"

    line 29
  4. Very specific:

    When a remediation is natural (for example pnpm deploy-all --quick for a stale sandbox manifest), ask the user.

    line 56
  5. Very specific:

    Run /agentic-community-college:list.

    line 80
  6. Very specific:

    Run /agentic-community-college:create-lesson against this course to author the first lesson.

    line 81
  7. Distinctive:

    Do not run git commands from this skill.

    line 84

≈ “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

Course Creator Skill

You are scaffolding a fresh ACC content plugin. The output is a new directory on disk that, once enabled in Claude Code, hosts one or more lessons authored later by lesson-creator.

ACC ships zero domain probes. Every prerequisite check a lesson in this course needs is declared here, in the course's plugin.json. Four probe kinds are supported.

Step 1: Target directory

Ask the user where the new course should live. Default suggestion: ~/workspace/acc/acc-<domain>/ (for example ~/workspace/acc/acc-walrus/). Refuse to overwrite an existing non-empty directory. If the target exists, ask for a different path.

mkdir -p <target>/lessons/. Initialize git: cd <target> && git init --initial-branch=main.

Step 2: Plugin identity

Collect:

  • Plugin name (the name field in plugin.json). Convention: acc-<domain> (for example acc-claude-sdk). Lowercase, kebab-case.
  • One-line description for the manifest.
  • Author: default to git config user.name, ask to confirm.
  • Keywords: comma-separated list (for example walrus, storage, sui, course).

Step 3: Probe declarations

Ask the user: "What prerequisites does any lesson in this course need?"

Common categories:

  • Tools on PATH: node, pnpm, sui, docker, language toolchains. Each becomes a shell-exit-zero probe that runs the binary with --version.
  • Sibling repos or fixture directories: each becomes a filesystem-exists probe.
  • Local services: RPC endpoints, dev servers. Each becomes an http-get probe with expected_status: 200, optionally with a shell remediation that brings the service up.
  • Other Claude Code plugins: each becomes a claude-plugin-enabled probe that reads ~/.claude/settings.json.

For each probe collect:

  • id: kebab-case, unique within this course (sandbox-manifest-reachable).
  • kind: one of the four kinds.
  • message_pass: one short sentence shown on success.
  • message_fail: shown on failure. Make it actionable: name the command, URL, or file to fix.
  • params: kind-specific, see the table.
  • remediation (optional): { kind: "shell", command, cwd?, timeout_ms? }.

Probe-kind cheat sheet

KindRequired paramsOptional paramsExample use
filesystem-existspath (absolute or ~/-prefixed)nonesibling repo, fixture dir
http-geturlexpected_status (default 200), timeout_ms (default 5000), expected_body_regexservice is alive, manifest reachable
shell-exit-zerocommandargs[], timeout_ms (default 10000), expected_stdout_regexsui --version, docker info
claude-plugin-enabledplugin_keynoneanother Claude Code plugin this course depends on

When a remediation is natural (for example pnpm deploy-all --quick for a stale sandbox manifest), ask the user. On yes, gather command, optional cwd (supports ~/), optional timeout_ms (default 60000).

Render each probe back to the user before adding it. Skip this step with the user's explicit confirmation: many courses have no prerequisites, and probes: [] is valid.

toolkit@contract-hero is optional. The conductor offers publish-html only when that plugin is enabled. Do not declare a probe for it.

Step 4: Emit files

Generate, in order:

  1. <target>/.claude-plugin/plugin.json from templates/plugin.json.tmpl: name, description, author, keywords, probes array. The template hardcodes "commands": ["./commands/start.md"] and accContent.lessons = "./lessons/".
  2. <target>/commands/start.md from templates/start.md.tmpl: plugin name plus a one-paragraph welcome the user supplies (or a sensible default). This is the learner's entry point, /<name>:start. It delegates to ACC's course-engine skill with a course filter pinned to this plugin.
  3. <target>/README.md from templates/README.md.tmpl.
  4. <target>/CLAUDE.md from templates/CLAUDE.md.tmpl.
  5. <target>/.gitignore from templates/.gitignore.tmpl (node_modules/, .vite/, dist/, .acc/, OS junk).
  6. <target>/lessons/.gitkeep.

Fill {{ key }} placeholders with a plain string replace. Variables: {{ name }}, {{ description }}, {{ description_short }}, {{ author }}, {{ keywords_json }}, {{ probes_json }}, {{ welcome_paragraph }}. {{ probes_json }} is the JSON array body (empty string when there are no probes).

Step 5: Next steps

Tell the user:

  1. Push the course repo to a remote and add it to a marketplace, or install it locally, then claude plugins install <name>@<marketplace> so ACC's discovery picks it up.
  2. Run /agentic-community-college:list. The new course appears under "Discovered courses" with an empty catalog.
  3. Run /agentic-community-college:create-lesson against this course to author the first lesson.
  4. git add -A && git commit -m "chore: scaffold <name>" inside the target dir.

Do not run git commands from this skill. Do not push to a remote unless the user asks.

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.