AgentStack
SKILL verified MIT Self-run

Skill Repo Maintainer

skill-coikexxx-skill-skill-repo-maintainer · by coikexxx

Maintain a multi-skill repository that stores several Codex/OpenAI skills under a shared `skills/` directory. Use when Codex needs to add a new skill, refactor a repo into per-skill folders, normalize metadata such as `SKILL.md` and `agents/openai.yaml`, refresh repository-level indexes, or prepare a skill for packaging and distribution.

No reviews yet
0 installs
3 views
0.0% view→install

Install

$ agentstack add skill-coikexxx-skill-skill-repo-maintainer

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Skill Repo Maintainer? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Skill Repo Maintainer

Keep the repository organized by skill folder first.

Core workflow

  1. Identify the target skill name and create or update skills//.
  2. Ensure the skill keeps its own SKILL.md as the source of truth.
  3. Add or refresh agents/openai.yaml when UI metadata is needed.
  4. Store reusable support material inside that skill only:
  • references/ for documentation loaded on demand
  • scripts/ for deterministic helpers
  • assets/ for templates or bundled files
  1. Update repository-level indexes or README files after changing the skill set.
  2. Treat dist/.skill as a release artifact, not the editing source.

Repository normalization rules

  • Prefer one top-level folder per skill under skills/.
  • Keep cross-skill documentation in shared repo files such as README.md or docs/INDEX.md.
  • Do not mix one skill's references, assets, or scripts into another skill's directory.
  • Keep skill names in lowercase hyphen-case.
  • Keep YAML frontmatter limited to name and description.
  • Make the description explain both capability and trigger conditions.

When adding a new skill

  • Pick a short hyphen-case folder name.
  • Write SKILL.md for the reusable workflow, not for one-off project notes.
  • Create only the resource subdirectories the skill truly needs.
  • Add an index entry that records source path, package path, and current status.
  • If packaging is not done yet, mark the skill as source-only instead of inventing a release artifact.

When refactoring an existing repo into a skill repo

  • Move each distinct capability into its own folder under skills/.
  • Rewrite the root README so it describes the repository as a catalog of skills.
  • Add a compact shared index that lists every skill and its release state.
  • Standardize metadata before worrying about packaging.
  • Preserve already-built artifacts in dist/, but treat them as outputs.

Release-readiness check

Before packaging a skill, confirm:

  • SKILL.md frontmatter is valid.
  • referenced files and folders actually exist.
  • agents/openai.yaml matches the current skill purpose.
  • repository indexes mention the skill and package status.
  • the skill can still be understood without unrelated repo context.

Load references as needed

  • Read references/repo-checklist.md for a compact maintenance checklist.
  • Read references/release-readiness.md before packaging or publishing a skill.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.