hermes-conductor-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@hermes-conductor-mcp Set up a workspace on Conductor and have Claude write a script to calculate Fibonacci numbers."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
hermes-conductor-mcp
MCP server that exposes Conductor Cloud as tools for Hermes Agent, so Hermes (running as your Telegram brain) can dispatch real coding work to Conductor's sandboxed developer agents.
Telegram <-> Hermes Agent (brain) <-> this MCP server <-> Conductor Cloud (dev agents)Tools exposed
list_projects— repos available to create a workspace fromcreate_workspace— spin up a sandboxed workspace + first session (agent:claude|codex|cursor|acp)get_workspace_status— poll workspace lifecycle (initializing→ready→ ...)archive_workspace— tear a workspace down when donesend_message— send an instruction/prompt into a coding sessionget_session_status— poll whether the agent isidle|working|errorget_session_messages— read the full transcriptcancel_session— stop the agent mid-task
Typical flow a Hermes conversation would drive: list_projects → create_workspace →
send_message (the task) → poll get_session_status until idle → get_session_messages
for the result → archive_workspace.
Related MCP server: hermes-claude-bridge
Setup
Get a Conductor API key (requires Conductor Pro): https://app.conductor.build/users/api-keys
Install and build:
npm install npm run buildRegister it in Hermes's
~/.hermes/config.yaml:mcp_servers: conductor: command: "node" args: ["/Users/rohanarora/Desktop/hermes-conductor-mcp/dist/index.js"] env: CONDUCTOR_API_KEY: "<your conductor api key>"Restart Hermes. It should discover tools prefixed
mcp_conductor_*(e.g.mcp_conductor_create_workspace).Make sure Hermes's Telegram gateway is already configured (BotFather token + your user ID) — see https://hermes-agent.nousresearch.com/docs/user-guide/messaging/telegram. That part is independent of this server.
Local dev
CONDUCTOR_API_KEY=sk-... npm run devRuns the server directly via tsx over stdio — pair with any MCP inspector/client to test tool
calls without going through Hermes.
Available Tools
8 toolsarchive_workspaceA
Archive (deactivate) a Conductor workspace once its work is done.
Inputs: workspaceId (required).
cancel_sessionC
Cancel/stop the agent's current operation in a session.
Inputs: sessionId (required).
create_workspaceA
Create a new Conductor workspace (a sandboxed coding-agent environment) and start a session in it. Provide either projectId (existing repo already connected to Conductor) or repositoryUrl.
Inputs: agent (required), name, model, branch, +4 more.
get_session_messagesC
Retrieve the full conversation transcript for a coding agent session.
Inputs: sessionId (required).
get_session_statusA
Poll whether a coding agent session is idle, working, or errored. Poll this after send_message to know when the agent finished.
Inputs: sessionId (required).
get_workspace_statusA
Check the lifecycle status of a Conductor workspace (initializing, ready, sleeping, archived, deleted, updating).
Inputs: workspaceId (required).
list_projectsB
List repositories available to create Conductor workspaces from.
Inputs: no named parameters.
send_messageA
Send a prompt/instruction to a coding agent session. Use the sessionId returned by create_workspace.
Inputs: message (required), sessionId (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 8 tools
Each tool targets a distinct resource and action: projects, workspaces, sessions, and operations on them. No two tools appear to overlap in purpose, making selection unambiguous.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., list_projects, create_workspace, send_message). This creates a predictable and readable interface.
With 8 tools, the server covers workspace lifecycle and session management without bloat. Each tool serves a clear need, and the count is well within the ideal range.
The surface covers workspace creation, status, archiving, session messaging, polling, transcripts, and cancellation. A notable gap is the lack of a list_workspaces tool to enumerate existing workspaces, but core workflows are otherwise complete.
Maintenance
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
Hosted MCP server and Agent Skills for specs, engineering threads, and project memory.
MCP server for building and testing AI agents with multi-model experimentation and insights.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceAn MCP server that allows Claude Desktop and other MCP clients to delegate tasks to a local Hermes Agent for automation and persistent operations.46 PyPI70Apache 2.0
- AlicenseNot gradedqualityCmaintenanceMCP server that enables Hermes Agent to delegate development tasks to Claude Code CLI, reusing the Claude Code subscription without additional API costs.1MIT
- AlicenseNot gradedqualityDmaintenanceA local MCP server that lets Hermes supervise Claude Code, delegating focused coding, research, or review tasks to the Claude Code CLI and managing worker sessions, background jobs, cancellations, and read-only reviews.3MIT
- FlicenseNot gradedqualityDmaintenanceThis MCP server exposes Claude Code as a tool, allowing MCP clients to run Claude Code agents in specified working directories. It manages session persistence and returns structured results including summary, changed files, tests, and blockers.-