Tendem Mcp
Delegate mixed AI-and-human work from your coding agent by creating and tracking Tendem tasks without leaving the IDE.
Overview
Tendem MCP is an MCP server for the Build phase that lets your agent create, manage, and track Tendem tasks backed by AI and human workers.
What is this MCP server?
- Create and manage Tendem tasks from Claude Code, Cursor, or other MCP clients via stdio
- Hybrid AI + human agent workflow for work that needs verification or specialist input
- Track task progress and outcomes from the agent session
- PyPI package tendem-mcp v0.1.9 with uvx runtime hint
- Requires TENDEM_API_KEY from agent.tendem.ai/tokens
- Server version 0.1.9
- Transport: stdio
- 1 required secret env var: TENDEM_API_KEY
What problem does it solve?
Agents can draft code fast but stall on work that needs humans, external verification, or durable task tracking outside the repo.
Who is it for?
Solo builders running agent workflows who regularly offload verification, ops, or specialist steps to a managed human-in-the-loop queue.
Skip if: Teams that only need local todo lists or GitHub Issues with no Toloka/Tendem account and no hybrid fulfillment.
What do I get? / Deliverables
After you register tendem-mcp with TENDEM_API_KEY, your agent can file and monitor hybrid tasks on Tendem from the same session where you ship code.
- Tendem tasks created and updated from agent tool calls
- Centralized tracking for AI-and-human fulfillment steps
- stdio MCP bridge at server version 0.1.9
Recommended MCP Servers
Journey fit
Task orchestration with human reviewers sits in Build under project management—the phase where solo builders coordinate what ships next. PM subphase covers backlogs, delegation, and status tracking that Tendem exposes through MCP tools.
How it compares
Hybrid task platform MCP integration, not a local project-management skill or a generic LLM wrapper.
Common Questions / FAQ
Who is Tendem MCP for?
Indie and solo builders using Claude Code, Cursor, or Codex who want MCP-driven create-and-track access to Tendem’s AI plus human task platform.
When should I use Tendem MCP?
Use it during Build and Operate when you need to assign real-world or human-reviewed tasks from the agent and follow status without leaving the IDE.
How do I add Tendem MCP to my agent?
Install tendem-mcp from PyPI, run it with uvx over stdio, and set the secret environment variable TENDEM_API_KEY from https://agent.tendem.ai/tokens in your MCP client config.