OpenSkillIndex← back to search
claude-code

Skilllibrary community-skill-harvester

Find external skills from public registries, GitHub repos, and official skill collections, then evaluate them for quality, licensing, and fitness for adoption. Use when looking for existing skills before building from scratch, evaluating external skill quality, or migrating community skills into a local library. Do not use when building a novel skill with no external precedent or for quick one-off evaluation (just read the skill directly).

install
source · Clone the upstream repo
git clone https://github.com/merceralex397-collab/skilllibrary
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/merceralex397-collab/skilllibrary "$T" && mkdir -p ~/.claude/skills && cp -r "$T/01-package-scaffolding/community-skill-harvester" ~/.claude/skills/merceralex397-collab-skilllibrary-community-skill-harvester && rm -rf "$T"
manifest: 01-package-scaffolding/community-skill-harvester/SKILL.md
source content

Community Skill Harvester

Harvests skills from public registries and evaluates them for adoption.

Procedure

1. Search for relevant skills

Search these sources in order:

# GitHub search by topic
gh search repos --topic agent-skills --limit 20 --json fullName,description,stargazersCount

# Anthropic skills repo
gh api repos/anthropics/skills/contents/skills --jq '.[].name' 2>/dev/null

# Direct GitHub code search
gh search code "name:" --filename SKILL.md --limit 20

2. Evaluate skill quality

For each candidate, apply this checklist:

CriterionCheckWeight
Has SKILL.mdFile existsRequired
Clear purposeNon-vague descriptionHigh
Concrete proceduresNumbered steps, not platitudesHigh
License specifiedSPDX identifier presentRequired
Active maintenanceCommits in last 6 monthsMedium
Usage evidenceStars, forks, downloadsMedium

Scoring: Required checks must pass or skill is rejected. Score 0-6 on remaining criteria. Score ≥4 = adopt candidate. Score 2-3 = adapt candidate. Score <2 = reject.

3. License compatibility check

Acceptable licenses for adoption: 

Apache-2.0
, 
MIT
, 
BSD-2-Clause
, 
BSD-3-Clause
, 
ISC
, 
CC0-1.0
, 
Unlicense
.

Copyleft licenses (

GPL-*
, 
AGPL-*
): flag for manual review before adoption.

No license: do not import.

4. Extract patterns

Before importing, document:

  • Structure patterns worth adopting
  • Anti-patterns to avoid
  • Adaptations needed for local conventions

5. Create import proposal

# Skill Import Proposal: [name]

## Source
- URL: [github.com/org/repo/skills/name]
- License: [SPDX]
- Quality Score: [N/6]

## Rationale
[Why adopt this skill]

## Required Adaptations
- [ ] [Adaptation 1]
- [ ] [Adaptation 2]

## Recommendation
[ADOPT | ADAPT | REFERENCE_ONLY | REJECT]

6. Execute import (if approved)

mkdir -p COPILOT/[skill-name]-COPILOT-L
# Copy and adapt SKILL.md
# Add provenance section documenting origin, license, import date, modifications
git add COPILOT/[skill-name]-*
git commit -m "feat: import [skill-name] from [source]"

Output contract

A harvest report containing:

  • Sources searched and count
  • Skills evaluated with scores
  • Import proposals for candidates scoring ≥2
  • Patterns discovered during evaluation

Failure handling

  • Registry unavailable: Fall back to direct GitHub search
  • License unclear: Do not import; flag for manual review
  • Skill format incompatible: Extract concepts, rewrite in local format
  • Quality too low: Document as "reference only", do not import

References