Skip to main content

Connect DeepWiki to Claude Code

Give Claude Code governed access to DeepWiki’s 4 tools through Runlayer, with identity, policy, scanning, and audit on every call.

Set up

Add DeepWiki to Claude Code in four steps

Admins approve the server once. Everyone in Claude Code then gets it through the same governed connection.

  1. 1

    Approve DeepWiki in Runlayer

    Add DeepWiki from the catalog and choose who can use it and which of its tools are allowed.

  2. 2

    Connect Claude Code to Runlayer

    Runlayer’s CLI installs the Runlayer Plugin natively in Claude Code, and an Anthropic org install also reaches Claude Code. With direct tool listing turned on for your deployment, DeepWiki’s 4 tools show up in Claude Code’s own tool list (supported Claude Code versions).

    Claude Code setup guide
  3. 3

    Sign in to DeepWiki

    Sign in the way DeepWiki requires. Runlayer holds the credentials so they stay out of the client’s configuration.

  4. 4

    Use DeepWiki from Claude Code

    DeepWiki’s approved tools appear in Claude Code. Every call is checked against policy and recorded.

Tools

4 DeepWiki tools in Claude Code

Runlayer policy decides which of these Claude Code users can call.

  • Ask question

    Ask any question about a GitHub repository and get an AI-powered, context-grounded response.

  • Ask wiki question

    Ask any question about a GitHub repository's codebase and get an AI-powered answer grounded in its DeepWiki.

  • Read wiki contents

    View documentation about a GitHub repository.

  • Read wiki structure

    Get a list of documentation topics for a GitHub repository.

Control

What Runlayer can see and stop in Claude Code

Some controls apply to every client at the gateway. Others depend on what Claude Code exposes to Runlayer.

  • Gateway policy (available)

    Every DeepWiki call from Claude Code passes through Runlayer, where your policies can allow, alert, mask, or block it.

  • Audit trail (available)

    Each call is recorded with the user, Claude Code as the client, the tool, and the policy and scanner decisions.

  • Client hooks (available)

    Capture supported prompts, reasoning, local and MCP tool calls, responses, and scanner decisions in real time.

  • Anthropic inference hooks (available)

    With Claude Enterprise, Anthropic sends each governed Claude Code prompt to Runlayer before inference, with nothing installed on devices (Anthropic beta).

  • OTLP telemetry (available)

    Claude Code can export OpenTelemetry logs to Runlayer for observe-only session reconstruction.

  • Stop a call before it runs (available)

    Runlayer can deny a tool call before Claude Code runs it.

  • Block output after it runs (available)

    Runlayer can block a tool’s output before Claude Code’s model sees it.

  • Redact output in place (available)

    Runlayer can mask sensitive parts of a tool’s output in Claude Code and pass the rest through.

In Claude Code

  • When a tool fails, Runlayer records the error but cannot hide it.

Frequently asked questions

Connect more

Find another combination