Skillpack maintainer

Skill pbans-agent/skillpack/skills/skillpack-maintainer

Git-backed source of truth for reusable AI Agent Skills

Install

npx -y skills add pbans-agent/skillpack --skill skillpack-maintainer

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Maintains a Git-backed Skill Pack repository of reusable Agent Skills. Use when the user asks to create, edit, deduplicate, validate, test, install, update, version, package, or improve skills in a central skill repository.

SKILL.md

3.8 KB, as published. Nobody here has run it

Skill Pack Maintainer

Purpose

Maintain this Skill Pack repo as the source of truth for reusable Agent Skills.

Location

The Skill Pack repo is at the path where you cloned it (default: ~/.skillpacks/metaportal-skillpack).

Core Commands

python3 scripts/skillpack.py list              # See all skills
python3 scripts/skillpack.py validate          # Validate everything
python3 scripts/skillpack.py install --scope personal --profile all   # Install
python3 scripts/skillpack.py status --scope personal                  # Check state
python3 scripts/skillpack.py update --scope personal --profile all    # Update

Workflow

When Creating a New Skill

Identify what the skill should do and when it should trigger.
Create skills/<skill-name>/ directory (lowercase, hyphens only).
Write SKILL.md with proper frontmatter (name, description, optional metadata).
Keep SKILL.md focused — move detailed content into references/.
Add scripts to scripts/ if needed (non-interactive, --help, structured output).
Add evals to evals/ if behavior should be testable.
Update skillpack.yaml if adding to specific profiles.
Run python3 scripts/skillpack.py validate.
Commit with: feat(<skill-name>): <description>.

When Editing an Existing Skill

Pull latest main: git pull --ff-only.
Create a focused branch: git checkout -b agent/update-<skill-name>-<purpose>.
Make minimal, focused edits.
Run python3 scripts/skillpack.py validate.
Run skill-specific evals if they exist.
Commit with: fix(<skill-name>): <description> or feat(<skill-name>): <description>.
Summarize: what changed, why, validation results, risks.

When Improving Install/Sync Tooling

Pull latest main.
Create branch: git checkout -b agent/tooling-<purpose>.
Edit scripts/skillpack.py or related files.
Run existing tests: python3 -m unittest discover -s tests -p 'test_*.py' -v.
Add new tests for new behavior.
Commit with: feat(tooling): <description>.

Guardrails

Do not overwrite ad-hoc/unmanaged skills during install.
Do not store credentials or private secrets in skills.
Do not add broad/destructive scripts without dry-run support.
Do not resolve Git conflicts blindly — inspect both versions.
Do not bloat SKILL.md when a referenced file is better.
Do not push directly to main without user approval.

SKILL.md Frontmatter

Required fields:

---
name: skill-name              # Must match directory name
description: >-
  Clear, trigger-rich description. Use when the user asks to...
---

Optional metadata:

metadata:
  skillpack.owner: metaportal-agents
  skillpack.version: "0.1.0"
  skillpack.tags: "tag1,tag2"
  skillpack.source: "metaportal-skillpack"

Validation Checks

The validate command checks:

skillpack.yaml exists and is parseable
Every skill has SKILL.md with valid frontmatter
Directory name matches frontmatter name
name is lowercase with hyphens only
description exists and is not too short or vague
Referenced files in markdown links exist
Scripts have help text
No secret-looking files

File References

Skill Authoring Standards — detailed rules for writing skills
Contributing Guide — agent contribution workflow