OSINT Benchmark Dashboard

Interactive explorer for canonical knowledge graph, episode traces, source platform records, and benchmark ranking.

Episode Explorer

Task ID:
Task Type:
Question
Ground Truth Answer:
Agent Answer:
Correct:

Graph Controls

Node Types

Graph Explorer

Layer: Canonical Graph
matched edge predicted only truth only

Node Inspector

Click a node to inspect attributes and neighbors.

Edge Inspector

Click an edge to inspect relation details.

Original Database Explorer

Selected Source Record

Click a row in the database table to inspect full JSON.

Benchmark Summary Radar

Episode Reward and Graph F1

Benchmark Leaderboard