Codex Claude Code MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CLAUDE_CODE_MCP_PATH | No | Absolute path to Claude Code executable. Mutually exclusive with CLAUDE_CODE_MCP_COMMAND. | |
| CLAUDE_CODE_MCP_COMMAND | No | Command name for Claude Code executable. Mutually exclusive with CLAUDE_CODE_MCP_PATH. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| claude_code_modelsA | List models available to the authenticated Claude Code CLI without running a model inference turn. Returns model identifiers, resolved models, descriptions, and supported effort or mode capabilities. |
| claude_code_runA | Run a new Claude Code session and keep this single MCP call open until it completes. Do not poll. For long tasks, a Codex parent should delegate this tool call to a Codex subagent and wait with Codex-native agent waiting. |
| claude_code_resumeA | Resume a persisted Claude Code session and keep this single MCP call open until the turn completes. Do not poll. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 3 tools
Each tool has a clearly distinct purpose: listing models, starting a new session, and resuming an existing session. There is no overlap in their functions, making it easy for an agent to select the correct one.
All tool names follow a consistent pattern with the prefix 'claude_code_' followed by a verb or noun that clearly indicates the action. This uniformity makes the tool set predictable and intuitive.
With only three tools, the server is tightly scoped to its purpose of managing Claude Code sessions. Each tool provides a distinct, necessary function, and the small count is appropriate for the narrow domain.
The core workflow of listing models, running a new session, and resuming a session is well covered. A minor gap is the lack of a session management tool for listing or stopping active sessions, but these may not be necessary given the synchronous nature of the run and resume tools.