Open-source context engine for physical assets
One asset context. Every decision.
Bring your operational systems together in one live, shared view — so people, software, and AI agents can find answers and take action without moving your data.
$ uv tool install henosis-engineOpen-source · self-hosted · MIT · Python 3.11+
Built around how asset decisions are made
henosis organises information about your operation into five clear roles:
Identity
What is it?
a well, a pump, a work order
Location
Where is it?
a wellhead at 27.7°S, 140.3°E
Observation
What's it doing, now and over time?
gas rate over the last 30 days
Documents
What do we know about it?
the P&ID, the last inspection report
Relationships
What's it connected to?
the pipeline it feeds, its open defects
Decision → Action
Act on the asset, then carry its new state into the next decision.
Context that stays currentThese roles give each piece of information a clear purpose. Locations can appear on maps, observations can be plotted over time, and relationships can be followed directly. The meaning is built into the model instead of recreated in every tool.
Reach the answer with less searching
A generic graph shows what is connected, but not what each connection means. henosis gives information a clear role and provides tools for querying it, so agents spend less time working out where to look.
The comparison below is illustrative, not a measured benchmark.
Generic knowledge graph
Explores the graph to discover what matters
Access pattern shown
Agent
Graph traversal
henosis
Uses typed roles to retrieve the relevant context
- Identity
- Location
- Observation
- Documents
- Relationships
Purpose-built interfaces
Human
Canvas
App
REST API
Agent
MCP tools
Why henosis
One model, three interfaces
People work in the Canvas, applications use the REST API, and AI agents call MCP tools. Each works from the same definitions and context.
MIT licensed and self-hosted
Run henosis in your own environment. The code is open under MIT, so you can inspect, adapt and extend it.
Read data where it lives
henosis queries historians, maintenance records, maps and documents in their source systems. There is no migration or duplicate data layer.
Write actions back to source systems
People and agents can act on what they find, with governed changes recorded in the systems where the work happens.
Works with what you already run
Connect the systems that already hold your operational context. henosis ships with connectors for historians, databases, files, geospatial services, APIs, and document stores.
Historian & signals
- AVEVA PI (OSIsoft)
- REST / HTTP APIs
Databases
- PostgreSQL
- MySQL
- Snowflake
- SQLite
Files & lakes
- Parquet
- CSV
Geospatial services
- ArcGIS
- WFS
Documents
- Document stores
Need another source? Register a connector without forking or changing the core.
See the model at work
See how the agent identifies a constrained well, traces the problem downstream, recommends an inspection and raises a work order. The recordings show the real henosis Canvas.
Good morning.
The demo uses synthetic data and public information from the South Australian PEPS-SA petroleum register and Geoscience Australia. It contains no operator-confidential data.
Bring the same context into Claude
This unedited Claude Desktop session starts with an open question, traces the issue through the Cooper Basin network, raises a work order and produces a shareable brief. Use the chapters to jump to any step.
Try henosis in three commands
Install the Python 3.11+ CLI, create the included Cooper Basin example, and start the local server.
$ uv tool install henosis-engine# Create the worked example
$ henosis init cooper-basin# Start the explorer, REST API and MCP server
$ henosis serve -p cooper-basin --mcpOpen http://127.0.0.1:8000 to explore the live model. Search across entity types, then follow an asset’s relationships, readings, location, and documents.
Find the full setup guide and model grammar on GitHub and in the docs.