Deadline MCP Server
Provides AI access to a Thinkbox Deadline 10 render farm via its Web Service REST API, exposing 18 tools for farm status, job management, and monitoring through MCP.
README
Deadline MCP Server
AI access to a Thinkbox Deadline 10 render farm via its Web Service REST API, exposed as Model Context Protocol tools. Works with Hermes, Claude, and any MCP-compatible client.
- 18 tools — 9 read + 9 write (see Tools)
- Self-contained — own venv, no host dependencies beyond Python 3.10+
- Env-configured — no secrets in the repo
Requirements
| Requirement | Detail |
|---|---|
| Python | 3.10+ |
| Deadline Web Service | deadlinewebservice running, default port 8081 |
| Network | The target machine must reach the Web Service (same host, LAN, or tailnet) |
| Credentials | A Deadline user account (username + password — not email) |
Install
git clone https://github.com/officefiction/deadline-mcp-server.git
cd deadline-mcp-server
./install.sh
install.sh creates a venv/, installs mcp + httpx, and self-tests the tool registry. Expected output ends with:
tools (18): deadline_complete_job, deadline_delete_job, ...
Version pin: the server targets the MCP 1.x Python SDK (
mcp>=1.28,<2). MCP 2.x movedFastMCP— do not upgrade past 2.x without testing.
Configure
cp .env.example .env
# edit .env:
# DEADLINE_WS_URL=http://localhost:8081 (or your host)
# DEADLINE_WS_USER=<deadline username>
# DEADLINE_WS_PASS=<deadline password>
chmod 600 .env
| Variable | Required | Default | Description |
|---|---|---|---|
DEADLINE_WS_URL |
no | http://localhost:8081 |
Base URL of the Deadline Web Service |
DEADLINE_WS_USER |
yes | — | Deadline user account (not email) |
DEADLINE_WS_PASS |
yes | — | Deadline account password |
DEADLINE_MCP_READONLY |
no | false |
true = read-only tools only (no job mutations) |
Register in Hermes
Add this block to ~/.hermes/config.yaml (path must match where you cloned):
mcp_servers:
deadline:
command: "/opt/deadline-mcp-server/venv/bin/python"
args: ["/opt/deadline-mcp-server/server.py"]
env:
DEADLINE_WS_URL: "${DEADLINE_WS_URL}"
DEADLINE_WS_USER: "${DEADLINE_WS_USER}"
DEADLINE_WS_PASS: "${DEADLINE_WS_PASS}"
enabled: true
timeout: 120
connect_timeout: 60
Then put the three DEADLINE_WS_* values in the profile's .env (chmod 600), restart the gateway, and verify:
hermes mcp test deadline # expect 18 tools discovered
Register in other MCP clients
Point any MCP client at the stdio server:
{
"command": "/opt/deadline-mcp-server/venv/bin/python",
"args": ["/opt/deadline-mcp-server/server.py"],
"env": {
"DEADLINE_WS_URL": "http://localhost:8081",
"DEADLINE_WS_USER": "...",
"DEADLINE_WS_PASS": "..."
}
}
Tools
Read (always available)
| Tool | Description |
|---|---|
deadline_farm_status |
Overall farm snapshot: workers, jobs by state, groups, pools |
deadline_list_jobs |
List jobs, filter by state / user |
deadline_get_job |
One job by ID |
deadline_job_tasks |
Tasks (chunks) of a job, per-task status |
deadline_job_reports |
Job reports / render logs |
deadline_list_workers |
Farm workers (slaves) with status + heartbeat |
deadline_list_users |
Repository users |
deadline_list_groups |
Worker groups |
deadline_list_pools |
Job pools |
Write (disabled with DEADLINE_MCP_READONLY=true)
| Tool | Description |
|---|---|
deadline_suspend_job |
Stop new tasks from starting |
deadline_resume_job |
Resume a suspended job |
deadline_requeue_job |
Re-run a job's tasks |
deadline_complete_job |
Mark a job completed |
deadline_fail_job |
Mark a job failed |
deadline_pend_job |
Hold a job until released |
deadline_release_pending_job |
Release a pended job to the queue |
deadline_set_job_priority |
Change a job's priority |
deadline_delete_job |
Delete a job |
Job states
| State | Meaning |
|---|---|
| 0 | Unknown |
| 1 | Active |
| 2 | Suspended |
| 3 | Completed |
| 4 | Failed |
| 5 | Deleted |
| 6 | Pending |
Architecture
MCP client (Hermes / Claude / ...)
│ stdio (JSON-RPC)
▼
server.py (FastMCP, mcp<2)
│ HTTP Basic auth
▼
Deadline Web Service (deadlinewebservice, :8081)
│ REST
▼
Deadline repository (render farm)
Key API facts (this server handles them automatically):
- Workers are served under
/api/slaves(not/api/workers) - Job state lives in the top-level
Statfield - Auth is a Deadline user account — the Web Service returns
401without it
Development
# Regenerate/refresh docs against the live tool registry
python3 scripts/deadline_mcp_docs.py --outline # (internal tooling, not in this repo)
License
Internal tooling for the Office Fiction project. MIT.
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.