AgentRecommender

① LLM Retrieval · ② Tool Retrieval · ③ Draft Agent Candidates by GPT· ④ EasyRec* Scoring TopK

New here? Read the AgentSelect Benchmark “Getting Started” page to understand the dataset, pipeline, and evaluation setup.
Open Getting Started →

Manual Agent Compare Edit two agents and click Compare (uses EasyRec* scoring)

Query
Options are taken from the current results: Stage ① Top10 LLMs and Stage ② Top10 Tools.
Agent A
score=—
LLM
Tools (multi-select)
Tip: Hold Ctrl / ⌘ to multi-select. Tools can be empty.
Agent B
score=—
LLM
Tools (multi-select)
Tip: Hold Ctrl / ⌘ to multi-select. Tools can be empty.