Engine

Core evaluation engine for Agent Control.

The engine is the runtime that evaluates controls, resolves selectors, and runs evaluators. It is used by the server and SDK to apply control logic consistently.

Responsibilities

• Evaluator Discovery: Auto-discover evaluators via Python entry points
• Selector Evaluation: Extract data from payloads using selector paths
• Evaluator Execution: Run evaluators against selected data
• Caching: Cache evaluator instances for performance

Evaluator Discovery

The engine provides the public API for evaluator discovery:

from agent_control_engine import discover_evaluators, list_evaluators

# Discover all evaluators (runs once, safe to call multiple times)
discover_evaluators()

# Get all available evaluators
evaluators = list_evaluators()  # Returns dict[str, EvaluatorClass]

# Access a specific evaluator
regex_evaluator = evaluators.get("regex")

Evaluators are discovered via the agent_control.evaluators entry point group. Discovery:

1. Scans all installed packages for the entry point
2. Loads each evaluator class
3. Checks is_available() to verify dependencies
4. Registers available evaluators

Key Functions

Function	Description
discover_evaluators()	Scan entry points and register evaluators
list_evaluators()	Get all registered evaluators (triggers discovery)
ensure_evaluators_discovered()	Ensure discovery has run
get_evaluator_instance(config)	Get cached evaluator instance
evaluate_control(control, payload)	Evaluate a single control

Source Code

View the complete engine source code and implementation:

Agent Control Engine