$ loading_
Control Claude Code sessions in tmux with fine-grained orchestration and efficient extraction.
Overall, this is an open-source MIT-licensed MCP tool with no declared credentials or fixed remote endpoints, so the supply-chain posture is reasonably good. The main concern is local tmux-based interactive code execution, which warrants caution but not high risk.
No keys, tokens, or environment variables are declared; based on the materials, there is no direct credential exposure or abuse risk.
No remote host endpoints are listed, and no external data transfer is described; the materials show no explicit egress path.
The tool explicitly controls interactive Claude Code sessions inside tmux, which means local process launch and code execution as a normal capability; guard against unintended commands in-session.
The materials only describe mid-session steering, interruption, and result extraction, without broader file or resource permissions; interactive sessions may still reach local context, so use least privilege.
This is a third-party but open-source, MIT-licensed, auditable repository; however, it has 0 stars and unknown maintenance status, so supply-chain maturity is limited and deserves extra verification.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "claude-code-mcp" yet — see the docs or source repo.
Use claude-code-mcp to connect to a Claude Code session in tmux: first ask it to draft a Node.js API; halfway through, inspect the current progress and inject a new instruction: “Switch to TypeScript, add error handling, and preserve the existing route structure,” then extract only the final code files and a change summary.
Returns the final steered code output plus a concise change summary and key artifacts.
Start a Claude Code session with claude-code-mcp to refactor a large repository; if the session keeps producing long analysis with no concrete edits, interrupt it immediately and send a new instruction: “Scan only the auth module and list the top 5 refactoring priorities with file paths.”
Stops ineffective execution and produces a focused list of high-priority refactoring recommendations.
Use claude-code-mcp to manage a long-running Claude Code session without returning the full history; extract only the current task status, modified file list, unresolved issues, and recommended next steps for an orchestrator agent.
Outputs a compact structured session summary that an orchestrator agent can continue from efficiently.
Let AI safely control tmux sessions, windows, and panes for terminal automation.
Query Claude Code transcript analytics for cost, safety, audit, and efficiency insights.
Manage Claude Code sessions with hot restart and context compaction commands.
Wrap Claude Code CLI as MCP tools for headless coding sessions and automation.
Connect Claude Code with Emacs for buffer, file, org, git, and diff tasks.
Control a running Emacs via MCP for editing and status checks.