About Mcpskills CLI
Mcpskills CLI is an MCP server published by dhanababum in the Developer Tools category: generate Agent Skills from MCP server tools. Connects via Streamable HTTP, discovers tools, and outputs a skill with schema docs and a call script in the language of your choice.
The repository has 16 stars and 3 forks, with the last commit 8 months ago. Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.
Install
git clone https://github.com/dhanababum/mcpskills-cliThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Ask AI
Ask AI about Mcpskills CLI
Powered by Claude · Grounded in docs
Security checks
Static checks derived from the repository yesterday. Heuristics on the source, not an audit.
- ✓README presentThe repository ships documentation.
- ✓License declaredUsable under a stated license.
- ✓Tests presentA test suite exists in the repository.
- !Dependencies pinnedNo lockfile. Installs can pull different code over time.
- ✓No dynamic code executionSource scan found no eval, dynamic Function, or shell execution.
- !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.
README
mcpskills-cli
Generate Agent Skills from MCP server tools. Connects via Streamable HTTP, discovers tools, and outputs a skill with schema docs and a call script in the language of your choice.
Why bake MCP into skills?
Skills are easy to understand, static, edit, and most coding and non-coding agents are adopting them. Traditional MCPs load all tools into context and pollute the agent; token penalty is also costly because of loading every tool. This is why we transform MCP into skills: skills reduce token consumption because the agent does not load all skills into context—it only loads them when necessary.
But why do we even need MCP?
MCP is a standardized protocol for AI agents to communicate. Because of this, any system can easily adopt it and create the necessary tools specific to their needs.
Install
pip install mcpskills-cli
# or for development:
pip install -e .
Usage
mcpskills-cli --url <MCP_SERVER_URL> --token <TOKEN> [--name <NAME>] [--output <DIR>] [--script <LANG>] [--multi-skills]
| Flag | Default | Description |
|---|---|---|
--url |
(required) | MCP server endpoint |
--token |
(required) | Bearer token |
--name |
from URL | Server name (skill dir + credentials key) |
--output |
~/.cursor/skills |
Skills output directory |
--script |
bash |
Call script language: bash, python, node, go, rust |
--multi-skills |
false |
Generate a separate skill for each tool |
Examples
# Generate single skill with all tools (default)
mcpskills-cli --url http://localhost:8027/mcp/abc123 --token mytoken --name my-db
# Generate separate skill for each tool
mcpskills-cli --url http://localhost:8027/mcp/abc123 --token mytoken --name my-db --multi-skills
# Generate with Python call script
mcpskills-cli --url http://localhost:8027/mcp/abc123 --token mytoken --name my-db --script python
# Generate with Node.js call script
mcpskills-cli --url http://localhost:8027/mcp/abc123 --token mytoken --name my-db --script node
Credentials
Stored in ~/.mcps/credentials (INI format, chmod 600). One section per server, updated automatically.
[my-db]
url = http://localhost:8027/mcp/abc123/
token = mytoken
Rotate tokens by editing the file directly; no need to regenerate skills.
Generated Output
Default Mode (Single Skill)
~/.cursor/skills/<server-name>/
SKILL.md # Documents all tools with parameters
scripts/
call.<ext> # Calls any tool: ./call.<ext> <tool_name> '{"key":"val"}'
Multi-Skills Mode (--multi-skills)
~/.cursor/skills/<server-name>-<tool-name-1>/
SKILL.md # Documents single tool
scripts/
call.<ext> # Calls tool: ./call.<ext> <tool_name> '{"key":"val"}'
~/.cursor/skills/<server-name>-<tool-name-2>/
SKILL.md
scripts/
call.<ext>
Requirements
- Python >= 3.10
fastmcp>= 2.3jinja2>= 3.1
README fetched from the repository yesterday. The original is authoritative.