Braintrust Tracing
Braintrust Tracing reconstructs user actions within a session to understand decision-making and improve UX insights.
Install on your platform
We auto-selected Claude Code based on this skill’s supported platforms.
Run in terminal (recommended)
claude mcp add braintrust-tracing npx -- -y @trustedskills/braintrust-tracing
Or manually add to ~/.claude/settings.json
{
"mcpServers": {
"braintrust-tracing": {
"command": "npx",
"args": [
"-y",
"@trustedskills/braintrust-tracing"
]
}
}
}Requires Claude Code (claude CLI). Run claude --version to verify your install.
About This Skill
The braintrust-tracing skill enables AI agents to generate detailed execution logs and performance metrics directly within the Braintrust analytics platform. It provides visibility into agent reasoning steps, tool usage, and latency without requiring external instrumentation code.
When to use it
- Debug complex multi-step workflows where standard logging is insufficient.
- Analyze cost efficiency by tracking token consumption per interaction.
- Monitor real-time performance bottlenecks during high-volume agent operations.
- Generate audit trails for compliance or post-mortem analysis of failed tasks.
Key capabilities
- Automatic ingestion of agent traces into the Braintrust dashboard.
- Structured logging of reasoning chains and tool calls.
- Real-time visualization of execution latency and success rates.
- Support for custom metadata tagging during trace generation.
Example prompts
- "Run this data extraction task and send a full trace to Braintrust for analysis."
- "Execute the code generation workflow, ensuring all steps are logged with timestamps."
- "Perform the customer support simulation and capture detailed metrics on response time and accuracy."
Tips & gotchas
Ensure your AI agent has write permissions to the configured Braintrust project ID before enabling this skill. Traces may increase storage costs if not filtered by session duration or error status in the dashboard settings.
Tags
TrustedSkills Verification
Unlike other registries that point to live repositories, TrustedSkills pins every skill to a verified commit hash. This protects you from malicious updates — what you install today is exactly what was reviewed and verified.
Security Audits
| Gen Agent Trust Hub | Pass |
| Socket | Pass |
| Snyk | Pass |
🌐 Community
Passed automated security scans.