simple-fetch
Basic MCP skill demonstrating CLI-based execution pattern for fetching URL content
When & Why to Use This Skill
This Claude skill provides a foundational Model Context Protocol (MCP) execution pattern for fetching web content via a CLI-based workflow. It serves as a lightweight, extensible template for developers to learn MCP integration, enabling the automated retrieval of webpage data and documentation to enhance agentic context.
Use Cases
- Developer Education: Use as a reference implementation to understand the 'CLI args → MCP call → Result' pattern for building custom Claude skills.
- Automated Documentation Fetching: Quickly retrieve technical documentation or README files from URLs to provide up-to-date context for coding and debugging tasks.
- Web Content Extraction: Fetch raw content from specific web pages for summarization, data cleaning, or knowledge base enrichment.
- Skill Boilerplate: Serve as a minimal starting point for developing more complex web automation or data scraping agents within the MCP ecosystem.
| name | simple-fetch |
|---|---|
| description | Basic MCP skill demonstrating CLI-based execution pattern for fetching URL content |
Simple Fetch Skill
When to Use This Skill
Use this Skill to:
- Learn the basic skill pattern
- Fetch content from a URL
- Use as a template for creating custom skills
This is a demonstration skill showing the minimal CLI-based pattern.
What This Skill Does
Demonstrates the core skill pattern:
- Accept CLI arguments (--url)
- Call an MCP tool (fetch__fetch)
- Return result
Instructions
When you need to fetch content from a URL, execute:
cd /home/khitomer/Projects/mcp-code-execution-enhanced
uv run python -m runtime.harness scripts/simple_fetch.py \
--url "https://example.com"
Parameters
--url: The URL to fetch (required)
Example Usage
# Fetch a webpage
uv run python -m runtime.harness scripts/simple_fetch.py \
--url "https://docs.example.com/api"
# Fetch documentation
uv run python -m runtime.harness scripts/simple_fetch.py \
--url "https://github.com/owner/repo/README.md"
Expected Output
The skill returns the fetched content and prints:
- Success message with byte count
- Or error message if fetch fails
MCP Servers Required
Configure a fetch-capable MCP server in mcp_config.json:
{
"mcpServers": {
"fetch": {
"type": "stdio",
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
}
Technical Notes
- Pattern: CLI args → MCP call → Result
- Token cost: ~110 tokens (discover + read + execute)
- Time: <10 seconds
- Immutable: Parameters via CLI, no file editing needed
This skill demonstrates the foundation for creating more complex MCP workflows.