Provides integration with Inductiveautomation's Ignition SCADA/MES platform through its REST API, enabling AI assistants to manage gateway operations, handle backups, monitor logs, manage modules and projects, and perform license activation tasks.
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., "@Ignition MCP Server show me the current gateway status"
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.
A local Streamable HTTP front for ign mcp serve,
plus composite workflow tools, ign:// resources, and runbook prompts.
ign is the source of truth: every ign verb is proxied unchanged (the full ign tool catalog, 91 tools in ign 1.2.0),
credentials and profiles are ign's, and destructive verbs still require confirm: true.
This server never talks to a gateway itself.
Local development only. There is no client authentication. Keep the default loopback bind (
127.0.0.1). Do not expose it on a network.
An agent inspects gateway health, reads batch conditions and traceability, and discusses hold readiness through Ignition MCP.
Open the recordings with playback controls and written summaries. Recorded against a fictional batch process; playback runs at 1.2×.
Related MCP server: ignition-mcp-server
ign >= 1.2.0 on PATH (or IGN_BIN=/path/to/ign) with a configured profile
Python 3.11+, uv
ign workspace push cannot be confirmed over MCP in ign 1.2.0 (the verb is not in
ign's guarded set), so there is no local-workspace push composite yet.
uv sync
uv run ign-mcp --profile uat # http://127.0.0.1:8765/mcp
uv run ign-mcp --stdio # same server over stdioAdd to Claude Code:
claude mcp add --transport http ign-http http://127.0.0.1:8765/mcpOr point a client at the included mcp.example.json.
Env | Flag | Default |
|
|
|
|
| ign's active profile |
|
|
|
|
|
|
Gateway credentials: see ign profile --help (keychain or IGNITION_TOKEN* env).
mcp_server.py, run_server.sh, and the Docker image are gone; run uv run ign-mcp.
The IGNITION_MCP_* settings are gone too. The gateway URL and credentials now live in
ign profile, and the only settings left are the four in the table above.
Run ign adopt --project ign-cli once per gateway to deploy the WebDev routes the tag
and script verbs need.
Old tool names map onto ign verbs:
Old | New |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Tool | Steps | Guarded |
| status, license_status, redundancy_status, gan_status, connections, modules, doctor → | no |
| project_diff → project_sync (all-changed) → project_diff; profile A→B promotion; refuses |
|
| rig_down, rig_up, wait_gateway, status |
|
| tags_browse (recursed in Python, at most | no |
| tags_alarms_active, filtered | no |
Results: {"ok", "steps": [{tool, ok, code}], ...}; on failure also step and error (the ign envelope).
ign://status, ign://profiles, ign://projects, ign://rig/status, ign://logs/tail{?n}
health_check, bring_up_rig, sync_project(project, profile_a, profile_b), triage_alarm(path)
uv sync --extra dev
uv run pytest
uv run ruff check . && uv run ruff format --check .Tests use tests/fake_ign.py; no gateway needed.
This server cannot be deployed
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Give your AI agents the tools to build, manage, and run automation workflows.
Remote MCP for RunComfy: ComfyUI deployments, hosted models, LoRA training. 31 tools.
Universal AI API Orchestrator — 1,554 tools, 96 services. One install.
