Put an AI agent on rails: mount a Burr state machine as an MCP server so the agent can only take the next allowed step, with every step recorded and replayable.
SRE incident-investigation finite state machine for LLM agents, served over MCP by Theodosia. The agent keeps the full Grafana toolset; the FSM gates the procedure (triage, diagnose, verify, conclude) and the audit trail, not the tools.
Read-only on-call incident-triage agent. Reads metrics, logs, client load, and feature flags through MCP, correlates them, and reports the root cause with a confidence bounded by the evidence. Built on Theodosia. Winner, Crusoe challenge (Hermes/Nemotron on Crusoe Cloud), DevNetwork [AI+ML] Hackathon 2026.
A change/deploy-gate MCP agent built with Theodosia: ordered gates, a health gate, an audit trail, and a call to a filesystem MCP server.
A support-triage MCP agent built with Theodosia: investigate before you decide, enforced by the state-machine graph.
A toy MCP agent built with Theodosia: a coffee-order state machine an LLM drives one enforced step at a time.
Add a description, image, and links to the theodosia topic page so that developers can more easily learn about it.
To associate your repository with the theodosia topic, visit your repo's landing page and select "manage topics."