Multi-Agent Delegation Testing
Tests whether multi-agent systems properly delegate tasks through a chain and maintain accountability at each step. Uses real separate LLM calls per agent — no simulation.
—
Total Tests
📋→📋
Task Handoff Integrity
Agent A delegates to B. Does info survive the handoff?
10 tests
2 agents per test
📋→📋→📋
Chain of Custody
A→B→C chain. Telephone-game degradation test.
10 tests
3 agents per test
🎯
Delegation Decision Quality
Does the orchestrator delegate the right tasks to the right specialists?
10 tests
5-7 specialists
—
Overall Score
—
Handoff Integrity
—
Chain of Custody
—
Decision Quality
Model: —
Tokens: —
Avg Time: —
Test Results
Leaderboard
Loading...