Claude Code

Quick Setup

The fastest way to add BioContext7 to Claude Code — a single command, no config files needed:

npx (recommended)

claude mcp add biocontext7 -- npx -y @biocontext7/mcp@latest

This auto-installs the @biocontext7/mcp package and registers it as an MCP server.

HTTP remote (zero install)

claude mcp add --transport http biocontext7 https://mcp.biocontext7.com/mcp

No local install required — connects directly to the hosted BioContext7 API.

Manual Configuration

If you prefer editing config files directly, add to ~/.claude/settings.json or project .mcp.json:

{
  "mcpServers": {
    "biocontext7": {
      "command": "npx",
      "args": ["-y", "@biocontext7/mcp@latest"]
    }
  }
}

Or for HTTP remote:

{
  "mcpServers": {
    "biocontext7": {
      "type": "http",
      "url": "https://mcp.biocontext7.com/mcp"
    }
  }
}

From source (Python)

For local development with the Python MCP server:

{
  "mcpServers": {
    "biocontext7": {
      "command": "python",
      "args": ["-m", "mcp_servers.biotools.server"],
      "cwd": "/path/to/biocontext7/packages/biocontext7"
    }
  }
}

Verify Installation

After adding BioContext7, verify the MCP tools are available by asking Claude Code:

"What tools should I use for RNA-seq alignment?"

Expected: BioContext7's resolve-library-id fires and returns tools like STAR, HISAT2, Salmon.

You can also test each tool directly:

Test	Expected Result
`resolve-library-id` with "scanpy"	Returns scanpy tool ID and metadata
`get-library-docs` with a tool ID	Returns documentation snippets

Available MCP Tools

Core Tools

Tool	Description
`resolve-library-id`	Search 25,000+ bioinformatics tools by name or keyword
`get-library-docs`	Fetch versioned documentation for a specific tool by ID

ToolUniverse Server

Tool	Description
`search_tooluniverse`	Search 1,288+ bioinformatics API tools
`run_tooluniverse_tool`	Execute API calls through ToolUniverse
`list_tooluniverse_databases`	List available database backends
`get_tooluniverse_tool`	Get detailed tool metadata

Blarify Server

Tool	Description
`analyze_pipeline_dag`	Analyze pipeline DAG structure
`check_pipeline_types`	Check EDAM type compatibility
`validate_pipeline`	Full pipeline validation
`apply_pipeline_fix`	Apply auto-fix suggestions

Usage Examples

Once configured, use natural language in Claude Code:

"Search for RNA-seq alignment tools"
→ resolve-library-id(query="RNA-seq alignment")

"Get documentation for STAR aligner"
→ get-library-docs(libraryId="/biotools/star")

"Suggest a pipeline for ChIP-seq peak calling"
→ suggest_pipeline(task="ChIP-seq peak calling")

"Validate my pipeline at ./pipeline.yaml"
→ validate_pipeline(pipeline_path="./pipeline.yaml")

Transport Options

BioContext7 supports two MCP transport modes:

Transport	Config	Use Case
stdio	Default	Local Claude Code usage via npx
HTTP	`--transport http`	Remote or shared access

For HTTP transport, start the server manually:

bc7 serve --transport http --port 3000

Then configure Claude Code with the HTTP endpoint:

{
  "mcpServers": {
    "biocontext7": {
      "url": "http://localhost:3000/mcp"
    }
  }
}

On this page