D

dsh-harness-mcp-server

Free

Expose DeepSeek Harness agent capabilities as an MCP server (brain=Hermes, arms=Harness)

Visit Website
Related by shared tags and keywordsRestorkApplied Research · Blue Hen REKrowoc
ai-agentcordisdeepseek-harnessdsh-pluginmcpaiai assistantautomation

Overview

dsh-harness-mcp-server is an MCP server that exposes DeepSeek Harness agent capabilities, allowing any MCP client (e.g., Hermes) to drive Harness to execute real coding tasks. It starts a real MCP server (StreamableHTTP) inside Harness and bridges Harness's core services — ctx.agents, ctx.agentPresets, ctx.tools — so an external "brain" can delegate real work to Harness's "arms".

The server provides tools such as echo, harness_list_tools, agent_run, task_inbox, and task_result. agent_run runs a task synchronously and returns a structured result, while task_inbox pushes a structured task (task + memory context + cwd) to an async queue, and task_result polls a queued task's structured result. Every task result is structured, including sessionId, assistantText, toolCalls, toolResults, changes, verification, and leftovers.

Installation can be done from npm or from source. The MCP server listens on 127.0.0.1:8090 (StreamableHTTP). By default the server binds to 127.0.0.1 only. It exposes unauthenticated remote code execution — do not bind it to 0.0.0.0 or expose it to the internet/LAN without adding authentication, TLS, and a reverse proxy first.

Positioning: This is best used as a fallback tool, not a daily driver. It is useful for context isolation (huge refactors that would blow the client's context) or parallel execution of unrelated tasks. The agent session is reused per cwd, which is roughly 15–20× cheaper than one-shot dsh headless. Bash runs sandboxed (workspace-write).

Pricing

Free

Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.

Visit Website

Prices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.

Use Cases

  • Teams use dsh-harness-mcp-server to let an external MCP client delegate real coding tasks to Harness.
  • Teams use it to run a task synchronously and get a structured result.
  • Teams use it to push a structured task to an async queue.
  • Teams use it to poll a queued task's structured result.
  • Teams use it for context isolation during huge refactors.
  • Teams use it for parallel execution of unrelated tasks.

Features

AI assistant

Natural language interface

Workflow automation

Reviews

No reviews yet. Be the first to share your experience!

Rankings & collections