Skip to content
Back to search

skillport

Bring Agent Skills to Any AI Agent and Coding Agent — via CLI or MCP. Manage once, serve anywhere.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/gotalab--skillport/badge.svg)](/listings/gotalab--skillport)

HTML

<a href="/listings/gotalab--skillport"><img src="/api/listings/gotalab--skillport/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Install

uv tool install skillport
# or: pip install skillport

Add Skills

# Add from GitHub
skillport add anthropics/skills skills

# Or use a custom skills directory
skillport --skills-dir .claude/skills add anthropics/skills skills

Validate

skillport validate
# ✓ All 5 skill(s) pass validation

---

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 329 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # ⚓ SkillPort
  2. <div align="center">
  3. **The SkillOps Toolkit for Agent Skills**
  4. SkillOps = Validate, manage, and deliver skills at scale.
  5. [![Python](https://img.shields.io/badge/Python-3.10+-blue)](https://python.org)
  6. [![MCP](https://img.shields.io/badge/MCP-Enabled-green)](https://modelcontextprotocol.io)
  7. [![License](https://img.shields.io/badge/License-MIT-yellow)](LICENSE)
  8. </div>
  9. ---
  10. ## Why SkillPort?
  11. | When you... | SkillPort... |
  12. |-------------|--------------|
  13. | Use a coding agent without native skill support | Serves via MCP or CLI |
  14. | Build your own AI agent | Provides MCP server, CLI, and Python library |
  15. | Have 50+ skills and need the right one fast | Search-first loading ([Tool Search Tool](https://www.anthropic.com/engineering/advanced-tool-use) pattern) |
  16. | Check skills before deployment | Validates against the spec in CI |
  17. | Manage skill metadata programmatically | Provides `meta` commands |
  18. | Find a skill on GitHub | Installs with `add <url>` |
  19. Fully compatible with the [Agent Skills specification](https://agentskills.io/specification).
  20. ---
  21. ## Features
  22. ### Validate
  23. Check skills against the [Agent Skills specification](https://agentskills.io/specification).
  24. ```bash
  25. skillport validate # Validate all skills
  26. skillport validate ./skills # Validate specific directory
  27. skillport validate --json # CI-friendly JSON output
  28. ```
  29. Catches missing fields, naming issues, and spec violations before they cause problems.
  30. ### Manage
  31. Full lifecycle management from any source.
  32. ```bash
  33. # Add from GitHub (shorthand)
  34. skillport add anthropics/skills skills
  35. # Add from GitHub (full URL)
  36. skillport add https://github.com/anthropics/skills/tree/main/skills
  37. # Add from local path or zip
  38. skillport add ./my-skills
  39. skillport add skills.zip
  40. # Update, list, remove
  41. skillport update # Update all from original sources
  42. skillport list # See installed skills
  43. skillport remove <skill-id> # Uninstall
  44. ```
  45. ### Metadata
  46. Update skill metadata without editing files manually. Useful for automation and keeping skills consistent across a team.
  47. ```bash
  48. skillport meta get <skill> <key> # Get metadata value
  49. skillport meta set <skill> <key> <val> # Set metadata value
  50. skillport meta unset <skill> <key> # Remove metadata key
  51. ```
  52. ### Serve
  53. MCP server for clients that don't natively support Agent Skills.
  54. Inspired by Anthropic's [Tool Search Tool](https://www.anthropic.com/engineering/advanced-tool-use) pattern — search first, load on demand:
  55. | Tool | Purpose |
  56. |------|---------|
  57. | `search_skills(query)` | Find skills by description (full-text search) |
  58. | `load_skill(skill_id)` | Get full instructions + path |
  59. **Why search matters:** With 50+ skills, loading all upfront consumes context and hurts accuracy. SkillPort loads metadata only (~100 tokens/skill), then full instructions on demand.
  60. Works with Cursor, Copilot, Windsurf, Cline, Codex, and any MCP-compatible client.
  61. ---
  62. ## Quick Start
  63. ### Install
  64. ```bash
  65. uv tool install skillport
  66. # or: pip install skillport
  67. ```
  68. ### Add Skills
  69. ```bash
  70. # Add from GitHub
  71. skillport add anthropics/skills skills
  72. # Or use a custom skills directory
  73. skillport --skills-dir .claude/skills add anthropics/skills skills
  74. ```
  75. ### Validate
  76. ```bash
  77. skillport validate
  78. # ✓ All 5 skill(s) pass validation
  79. ```
  80. ---
  81. ## Connect to Agents
  82. Choose how to deliver skills to your AI agents:
  83. | Mode | Best for | Setup |
  84. |------|----------|-------|
  85. | [**CLI Mode**](#cli-mode) | Agents with shell access (Cursor, Windsurf, Codex, etc.) | Per-project |
  86. | [**MCP Mode**](#mcp-mode) | MCP-compatible clients, multi-project | One-time |
  87. ### CLI Mode
  88. For agents that can run shell commands. No MCP configuration required.
  89. ```bash
  90. skillport init # Initialize project
  91. skillport doc # Generate AGENTS.md with skill table
  92. skillport show <id> # Load full instructions for a skill
  93. ```
  94. How it works:
  95. 1. `skillport doc` generates a skill table in AGENTS.md
  96. 2. The agent reads AGENTS.md to discover available skills
  97. 3. When needed, the agent runs `skillport show <id>` to load full instructions
  98. ### MCP Mode
  99. For MCP-compatible clients. Install the server:
  100. ```bash
  101. uv tool install skillport-mcp
  102. ```
  103. Add to your client's config:
  104. ```json
  105. {
  106. "mcpServers": {
  107. "skillport": {
  108. "command": "uvx",
  109. "args": ["skillport-mcp"],
  110. "env": { "SKILLPORT_SKILLS_DIR": "~/.skillport/skills" }
  111. }
  112. }
  113. }
  114. ```
  115. <details>
  116. <summary>One-click install for popular clients</summary>
  117. **Cursor**
  118. [![Install MCP Server](https://cursor.com/deeplink/mcp-install-dark.svg)](cursor://anysphere.cursor-deeplink/mcp/install?name=skillport&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJza2lsbHBvcnQtbWNwIl0sImVudiI6eyJTS0lMTFBPUlRfU0tJTExTX0RJUiI6In4vLnNraWxscG9ydC9za2lsbHMifX0=)
  119. **VS Code / GitHub Copilot**
  120. [![Install in VS Code](https://img.shields.io/badge/VS_Code-Install_MCP_Server-007ACC?logo=visualstudiocode)](https://insiders.vscode.dev/redirect/mcp/install?name=skillport&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22skillport-mcp%22%5D%2C%22env%22%3A%7B%22SKILLPORT_SKILLS_DIR%22%3A%22~/.skillport/skills%22%7D%7D)
  121. **Kiro**
  122. [![Add to Kiro](https://kiro.dev/images/add-to-kiro.svg)](https://kiro.dev/launch/mcp/add?name=skillport&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22skillport-mcp%22%5D%2C%22env%22%3A%7B%22SKILLPORT_SKILLS_DIR%22%3A%22~/.skillport/skills%22%7D%2C%22disabled%22%3Afalse%2C%22autoApprove%22%3A%5B%5D%7D)
  123. **CLI Agents**
  124. ```bash
  125. # Codex
  126. codex mcp add skillport -- uvx skillport-mcp
  127. # Claude Code
  128. claude mcp add skillport -- uvx skillport-mcp
  129. ```
  130. </details>
  131. ---
  132. ## Organization
  133. Organize skills with categories and tags. Works with both CLI and MCP modes.
  134. ### Categories & Tags
  135. Use metadata to organize and filter skills:
  136. ```yaml
  137. # SKILL.md frontmatter
  138. metadata:
  139. skillport:
  140. category: development
  141. tags: [testing, quality]
  142. alwaysApply: true # Always available (Core Skills)
  143. ```
  144. ### Per-Client Filtering
  145. Expose different skills to different agents:
  146. ```json
  147. {
  148. "mcpServers": {
  149. "skillport-dev": {
  150. "command": "uvx",
  151. "args": ["skillport-mcp"],
  152. "env": { "SKILLPORT_ENABLED_CATEGORIES": "development,testing" }
  153. },
  154. "skillport-writing": {
  155. "command": "uvx",
  156. "args": ["skillport-mcp"],
  157. "env": { "SKILLPORT_ENABLED_CATEGORIES": "writing,research" }
  158. }
  159. }
  160. }
  161. ```
  162. ---
  163. ## Configuration
  164. | Variable | Description | Default |
  165. |----------|-------------|---------|
  166. | `SKILLPORT_SKILLS_DIR` | Skills directory | `~/.skillport/skills` |
  167. | `SKILLPORT_ENABLED_CATEGORIES` | Filter by category | all |
  168. | `SKILLPORT_ENABLED_SKILLS` | Filter by skill ID | all |
  169. | `SKILLPORT_ENABLED_NAMESPACES` | Filter by namespace | all |
  170. | `SKILLPORT_CORE_SKILLS_MODE` | Core Skills behavior (`auto`/`explicit`/`none`) | `auto` |
  171. ---
  172. ## Creating Skills
  173. Create a `SKILL.md` with YAML frontmatter:
  174. ```markdown
  175. ---
  176. name: my-skill
  177. description: What this skill does
  178. metadata:
  179. skillport:
  180. category: development
  181. tags: [example]
  182. ---
  183. # My Skill
  184. Instructions for the AI agent.
  185. ```
  186. See [Creating Skills Guide](guide/creating-skills.md) for best practices.
  187. ---
  188. ## Skill Sources
  189. | Source | Features | Target | URL |
  190. |--------|----------|--------|-----|
  191. | Anthropic Official | Document skills (docx, pdf, pptx, xlsx), design, MCP builder | All users | [GitHub](https://github.com/anthropics/skills/tree/main/skills) |
  192. | Awesome Claude Skills | Curated community collection, 2.5k+ stars | Discovery | [GitHub](https://github.com/ComposioHQ/awesome-claude-skills) |
  193. | Hugging Face Skills | Dataset creation, model evaluation, LLM training, paper publishing | ML/AI engineers | [GitHub](https://github.com/huggingface/skills) |
  194. | Claude Scientific Skills | 128+ scientific skills (bio, chem, ML), 26+ databases | Researchers | [GitHub](https://github.com/K-Dense-AI/claude-scientific-skills) |
  195. | ClaudeKit Skills | 30+ skills, auth, multimodal, problem-solving frameworks | Full-stack devs | [GitHub](https://github.com/mrgoonie/claudekit-skills) |
  196. | Superpowers | TDD, debugging, parallel agents, code review workflows | Quality-focused devs | [GitHub](https://github.com/obra/superpowers) |
  197. | Kubernetes Operations | K8s deployment, monitoring, troubleshooting | DevOps/SRE | [GitHub](https://github.com/wshobson/agents/tree/main/plugins/kubernetes-operations/skills) |
  198. ---
  199. ## Learn More
  200. - [Configuration Guide](guide/configuration.md)
  201. - [Creating Skills](guide/creating-skills.md)
  202. - [CLI Reference](guide/cli.md)
  203. - [Design Philosophy](guide/philosophy.md)
  204. ---
  205. ## Development
  206. > **Status:** Work in progress. APIs may change.
  207. ```bash
  208. git clone https://github.com/gotalab/skillport.git
  209. cd skillport
  210. uv sync
  211. # Run MCP server
  212. SKILLPORT_SKILLS_DIR=.skills uv run skillport-mcp
  213. # Run CLI
  214. uv run skillport --help
  215. # Run tests
  216. uv run pytest
  217. ```
  218. ---
  219. ## License
  220. MIT
  221. ## Local publishing credentials
  222. Local PyPI and TestPyPI tokens are read from 1Password only by `uv publish`,
  223. including its dry-run mode; builds receive no publishing credential.
  224. Provision the ignored `.env.publish-pypi` and `.env.publish-testpypi` maps with
  225. `UV_PUBLISH_TOKEN=op://<vault>/<item>/credential`; store one key per item under the `publish` tag.
  226. Use `make publish-check` or `make publish-test-check` to build all three distributions and run
  227. `uv publish --dry-run` without uploading. A dry-run reads the credential but does not
  228. prove its validity or upload permissions. `make publish` and `make publish-test` perform the uploads
  229. when explicitly requested. These local tasks require 1Password CLI and a current uv with publish dry-run support.
  230. The GitHub Actions release workflow continues to use PyPI Trusted Publishing (OIDC).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.