skillscatalog/registry
7 skills7 installs7 starsGitHub
Install
npx skills add https://github.com/skillscatalog/registrySkills in this repo
1Skill Createskill-create is a Skill Development generator for solo and indie builders who package procedural knowledge as Agent Skills instead of repeating setup in chat. It uses Python scaffolding (create_skill.py) to spin up a properly structured skill directory: SKILL.md, optional scripts folder, and manifest-friendly layout aligned with agentskills.io conventions. You reach for it in the build phase when you have identified a repeatable workflow—deploy checks, doc generation, review rituals—and want a consistent starting point before you flesh out triggers and body content. Templates cover bare documentation-only skills, skills with executable helpers, and skills that reference assets, so you do not hand-roll folder conventions on every new capability. The flow fits Cursor, Claude Code, Codex, and similar agents that consume SKILL.md from a repo. It does not replace authoring the actual procedure; it removes boilerplate so you can focus on invokeWhen rules, highlights, and validation. Pair it with catalog or review skills after the scaffold exists.1installs2Skill InstallerSkill Installer is a meta utility skill for solo builders who treat agent skills like dependencies: you want a verified bundle copied into the right folder for Claude Code, Cursor, Goose, or OpenCode without hand-moving files each time. The package ships a manifest that records SKILL.md and install_skill.py hashes, plus a Python installer that resolves shared config from the skills tree. Use it at the start of a build session when you adopt a new skill from a registry, after you clone a skills monorepo, or when you reset a laptop and need harness paths consistent again. It is deliberately not a marketplace browser—pair it with Prism discovery and then run install for the slug you chose. Expect filesystem and shell side effects; read the script before running in a repo with secrets. MIT-licensed skill metadata in the bundled manifest helps you audit what actually landed on disk.1installs3Skill Manifest GeneratorSkill Manifest Generator is a packaging skill for solo builders who publish or maintain Agent Skills in registries like Prism. It drives a Python workflow that scans skill folders, hashes files with SHA256, records sizes and types, and writes a manifest aligned with the agentskills.io manifest schema—including generator version and license SPDX metadata. Use it when you need Spec 20-style integrity proofs before users install your skill, or when automating catalog submissions. PyYAML improves frontmatter parsing but the script can degrade gracefully without it. The output makes downstream verification and supply-chain review practical for one-person skill authors who cannot manually checksum every release.1installs4Skill Publisherskill-publisher is a meta skill for solo builders who maintain custom agent skills and want them listed on Skills Catalog without hand-rolling upload steps each time. The package ships a manifest that fingerprints SKILL.md and a Python publish script, following the agentskills.io manifest schema. Workflow points you to skill keys settings and the my-skills console on skillscatalog.ai so you can confirm publication status after a run. It fits the Build phase when you are packaging procedural knowledge for Claude Code, Cursor, or Codex and need a consistent publish path from the repo. Expect to provide a valid skill key, a complete skill folder layout, and network access for the upload API. Outputs are a published listing consumers can discover through Prism-style directories and your own documentation links.1installs5Skill Safety Scannerskill-safety-scanner is a local checker skill for solo builders curating skills from registries who want web-parity safety grading without uploading code. It ships a Python safety_scan script that walks Agent Skill packages, flags hardcoded secrets and high-risk code patterns, and summarizes implied permissions such as filesystem, network, and subprocess use. Results are meant to align with skillscatalog.ai scanner grades so you can compare local runs with catalog audit panels. The package includes a versioned manifest with file hashes for SKILL.md and the scanner script, supporting reproducible reviews before you add a skill to Claude Code, Cursor, or Codex. Use it when evaluating unfamiliar skills from skills.sh-style sources, before CI adoption, or when you need a repeatable pre-install ritual alongside reading upstream SKILL.md triggers.1installs6Skill Searchskill-search is an agent skill that wires your coding agent to the skillscatalog.ai catalog through a small Python search script and a versioned manifest. Solo builders assembling Claude Code, Cursor, or Codex workflows often stall on “which skill already solves this?”—especially when juggling validate, build, and ship tasks in one repo. Instead of tab-hopping the registry, you invoke structured catalog queries from the agent session and get candidates you can install or compare. The package ships as a manifest-first SKILL.md plus scripts/search_catalog.py, with SHA256 integrity metadata suitable for teams that pin agent dependencies. You will need a catalog API key from the provider settings when authenticated search is required. Use it at the start of a project to map the landscape and again whenever you hit a new phase problem—testing, SEO, infra—that might already have a packaged skill. It complements Prism-style journey browsing by automating lookup against one specific registry backend.1installs7Skill Validatorskill-validator is a Skill Development checker for solo and indie builders who ship agent skills to registries like Prism or skills.sh. It runs a Python CLI against a skill folder and applies the same specification rules the catalog uses, so you catch frontmatter, structure, and compliance issues before users install your package. Use it when you are authoring or updating SKILL.md, bundling scripts, or preparing a release that must pass manifest and integrity expectations. The skill bundles validate_skill.py with argparse-driven usage, JSON reporting for automation, and a strict mode for tighter gates. It fits the Build phase on the agent-tooling shelf and also supports Ship-phase review when you treat spec compliance as a launch checklist. You get repeatable validation instead of guessing whether an agent runtime will accept your skill layout.1installs