Runtime AIBOM Emitter

Generates AIBOM entries in real-time by observing live agent runs. Connects to Episode Store and Gateway to discover what models, tools, and data sources an agent actually uses during execution.

Features

• Real-time emission tracking during agent execution
• Episode stream observation and parsing
• Integration with AIBOM Policy Engine
• Model/tool/data source usage discovery
• Emission filtering and summarization
• JSON export for audit trails

Quick Start

pip install -e .
python -m app.server

API runs on http://localhost:8700/v1

Emission Types

• MODEL_USED: Agent invoked an LLM model
• TOOL_INVOKED: Agent called a tool/function
• DATA_ACCESSED: Agent queried a data source
• POLICY_APPLIED: A policy was enforced

API Endpoints

Method	Endpoint	Purpose
GET	/v1/health	Health check
POST	/v1/emit	Record emission
POST	/v1/observe	Observe episodes
GET	/v1/summary/{agent_id}	Get agent summary
GET	/v1/emissions	List emissions
POST	/v1/publish	Publish to AIBOM
POST	/v1/export	Export as JSON

Integration

Connect to Episode Store for live observation:

Episode Store (port 8000)
        ↓
    Observer
        ↓
  Emissions
        ↓
AIBOM Engine (port 8600)

Testing

pytest tests/ -v

License

MIT