Ori
Offline-capable agentic IoT runtime that reasons about sensor data and acts under tiered authority.
Tested on Linux
Use it
Linux x86_64
You need first
- Runtimes:
python>=3.11
clone ori-runtime source
rm -rf ori-src && git clone https://github.com/ori-platform/ori-runtime.git ori-srcpip install editable with runtime and dev extras
cd ori-src && python3 -m pip install -e ".[runtime,dev]"install missing aiomqtt needed by tests
python3 -m pip install aiomqtt
Where the README and reality differ
- README: python -m pip install "ori-runtime[eval] @ git+https://github.com/ori-platform/ori-runtime.git@<commit-or-tag>" with literal <commit-or-tag> placeholder
What worked: use concrete tag/checkout or clone then pip install -e ".[runtime,dev]"; candidate 2 was used - README: python -m pip install -e ".[runtime,dev]" then pytest tests/ -v works
What worked: python3 -m pip install aiomqtt before pytest
Check that it works
python -m ori.runtime --helpThe output should contain “Ori Runtime”.
First thing to try
cd ori-src && python -m pytest tests/test_rule_engine.py -vWhat it does
These come from the README; pickfit didn’t check them one by one.
- Reason about sensor data in tiersNot tested
- Act on equipment under graded authority levelsNot tested
- Keep working without an internet connectionNot tested
- Keep a record of actions you can reconstructNot tested
How far pickfit got
Linux x86_64
- Installs
- Starts
- Main function works
- Works offline
macOS: the project says it runs there; pickfit hasn’t tried it.
Good for
- switch to inverter power on grid instability
- propose opening safety circuit on critical fault
- monitor PC system health on a laptop via psutil
An alternative to Grafana and AWS IoT Core.
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
Its main function works
- Replayed from scratch
- Works offline
- Checked on 30 Oct 2026
- Version 2.5.0
Differences the test machine hit
- README: pytest tests/ -v as full-suite smoke test
In the test: run pytest tests/test_rule_engine.py -v (143 passed) or scripts/pc_health_report.py; full suite did not finish in 600s (killed at ~77%) - README: bash scripts/smoke-runtime-local.sh
In the test: scripts/smoke-runtime-local.sh requires a .venv directory and a local-SLM config; verified with system python via scripts/pc_health_report.py instead