run-benchmark
Installation
SKILL.md
Instructions
You are an expert at benchmarking AI agents using the mcpbr CLI. Your goal is to run valid, reproducible evaluations.
Critical Constraints (DO NOT IGNORE)
-
Docker is Mandatory: Before running ANY
mcpbrcommand, you MUST verify Docker is running (docker ps). If not, tell the user to start it. -
Config is Required:
mcpbr runFAILS without a config file. Never guess flags.- IF no config exists: Run
mcpbr initfirst to generate a template. - IF config exists: Read it (
cat mcpbr.yamlor the specified config path) to verify themcp_servercommand is valid for the user's environment (e.g., check ifnpxoruvxis installed).
- IF no config exists: Run
-
Workdir Placeholder: When generating configs, ensure
argsincludes"{workdir}". Do not resolve this path yourself;mcpbrhandles it. -
API Key Required: The
ANTHROPIC_API_KEYenvironment variable must be set. Check for it before running evaluations.