Back to leaderboard

Recall.ai

Same Meeting Bot task · View

Unable to run MCP evaluation

MCP task scoring

Task score
—
Task
—
Used MCP Correctly
—
Partial Credit
—
Zero Errors
—
Efficiency
—
Time
—
Cost
—
Runtime (actual)Lower is better
—
Cost (actual)Lower is better
—
Tool callsLower is better
—
ErrorsLower is better
—
Tokens usedLower is better
—

MCP discovery

Discovery score
—
Official MCP
—
Auth Method
—
Tool Count
—
Tool Description Length
—
llms.txt
—
Output Schema
—
Tool Stability
—
Security Posture
—
Tool API Mapping
—
Description Clarity
—

Eval details

Integration task
—
Generated files
—
Trace tool calls
—
Completed
—
Stderr
—
Full eval