agentleFS
KubementatKubementat/ai_agent_skills/skills/skill-loader/SKILL.md

skill-loader

Skill2 starsChanged 8 months agoMITView on GitHub

Use when discovering skills from custom local skill collections stored in specific directories. Activates on /skill-loader, load skills, discover skills, or skill collection queries.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

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

    Run script via python3 scripts/extract_skill_metadata.py <group-name>

    line 117
  2. Very specific:

    Don't manually parse SKILL.md files - let the script handle YAML extraction

    line 118

≈ “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

/skill-loader — Discover Local Skills

Triggering Conditions

Use when:

  • You need to discover what skills are available in a local collection before selecting which to load for current task
  • User asks about available skills, skill collections, or wants to explore custom skill directories
  • Query contains /skill-loader, "load skills", "discover skills", or "available skills"
  • You're working with custom skill collections stored in specific directories (superpowers, ai-agent-skills, obsidian-skills)

NOT for:

  • Reading individual SKILL.md files directly - let the script handle discovery
  • Manual directory listing without YAML frontmatter parsing
  • Skills outside your configured local collection directories

Trigger

User invokes /skill-loader followed by a skill group name:

/skill-loader superpowers
/skill-loader ai-agent-skills
/skill-loader obsidian-skills

Or naturally without the prefix:

Load skills from the superpowers collection
What skills are available in ai-agent-skills?
Discover what's in my local skill collections

The script looks up the skill group in index.yaml and extracts metadata from all SKILL.md files automatically, outputting results in JSON format by default.

How to Use

Direct Command

python3 scripts/extract_skill_metadata.py <skill-group-name>

Example Output (JSON)

The script outputs available skills in JSON format by default:

[
  {
    "name": "superpowers-tdd",
    "description": "Use when implementing test-driven development workflows",
    "location": "/home/user/skills/superpowers/skills/tdd/SKILL.md",
    "original_name": "tdd"
  },
  {
    "name": "superpowers-planning",
    "description": "Use for systematic project planning and task breakdown",
    "location": "/home/user/skills/superpowers/skills/planning/SKILL.md",
    "original_name": "planning"
  }
]

Skill Discovery

Activation keywords:

  • /skill-loader followed by a skill group name
  • load skills, discover skills, available skills

Discipline Enforcement Table

ExcuseReality
"I'll just list the directory"Misses YAML frontmatter parsing, metadata extraction logic
"Just read SKILL.md files manually"Skips path expansion (~ → home), prefix handling, format selection (--format=xml)
"Agent discovers it naturally"Without skill, agent may not know to run Python script at all
"Script is just a helper"Script handles edge cases: missing index.yaml, empty directories, permission errors
"I can grep for SKILL.md files"Skips YAML parsing, name/description extraction, location tracking

Core principle: The script automates what would otherwise be tedious manual discovery. Trust the mechanism.

Common Excuses (and Why They're Wrong)

ExcuseReality
"I'll skip this and just list skills manually"Manual listing misses YAML frontmatter, returns inconsistent format
"The script is optional overhead"Script handles ~ expansion, error handling, format selection automatically
"Agent can figure it out without the skill"Agent needs to know to run python3 scripts/extract_skill_metadata.py

Pressure Scenarios

Scenario: Agent forgets /skill-loader prefix

Input: "Load skills from superpowers" Expected: Script still runs via extract_skill_metadata.py superpowers

Scenario: index.yaml missing or corrupted

Agent should see clear error message with available groups listed.

Scenario: No SKILL.md files in directory

Output empty JSON array [] instead of failing silently.

Rules

  • Output discovered skills in JSON format by default
  • Skip the full skill list output to user-facing chat if too verbose (14+ items)
  • Run script via python3 scripts/extract_skill_metadata.py <group-name>
  • Don't manually parse SKILL.md files - let the script handle YAML extraction

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.