showrunner
A tiny always-on server that coordinates multiple coding-agent sessions on one project (a 'show'), using a worker/director pattern with polling, leases, and SQLite state.
README
showrunner
A tiny always-on MCP server that coordinates multiple coding-agent sessions on one project (a "show"). Deploy it once. Then, from inside any project repo, tell any agent session, local or cloud, Claude Code or Cursor, "you're a showrunner worker" and it registers, starts pulling tasks, and reports back; tell a second session "you're the showrunner director" and it takes over planning. State lives on the server, not in any session, so sessions are cattle: kill one anytime and a new one picks up exactly where it left off. One SQLite file, one static dashboard, ten MCP tools.

(Reproducible: scripts/seed-demo.mts fills a local server with this demo show.)
A worker session, after the sentence:
> You're a showrunner worker.
⏺ showrunner:register({show: "mygame", kind: "claude-local"})
← member_id "eager-crane"
⏺ showrunner:await_work({member_id: "eager-crane"})
← task t-19ebd84582 "Invite links expire after 24h"
⏺ …works the task on a fresh branch, heartbeating update_task…
⏺ showrunner:update_task({status: "input-required", note: "renew flow or plain 410?"})
⏺ showrunner:await_work(…) ← the director's answer arrives here
Quickstart
fly launch --no-deploy # creates the app from fly.toml, skips first deploy
export SHOWRUNNER_TOKEN=$(openssl rand -hex 24) # keep this shell open
fly secrets set SHOWRUNNER_TOKEN=$SHOWRUNNER_TOKEN
fly deploy
Connect a local Claude Code session (same shell, so $SHOWRUNNER_TOKEN is still set):
claude mcp add --transport http showrunner https://<your-app>.fly.dev/mcp \
--header "Authorization: Bearer $SHOWRUNNER_TOKEN"
For Cursor, see the .cursor/mcp.json snippet in examples/
(every client's setup is in docs/OPERATING.md).
Now open a session in any project repo and say:
"You're a showrunner worker."
Open a second one and say:
"You're the showrunner director."
See DESIGN.md for why it's built this way, and docs/OPERATING.md for everything operational: client setup, the callboard tour, shared notes, env knobs, the CLI, verifying a deployment, FAQ, and security posture.
License
MIT, see LICENSE.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.