braintrust

Analyze and inspect Ara service traces and logs with SQL-like queries.

Updated Apr 17, 2026

Search Tags:#debugging #logs #cli #trace #instrumentation #braintrust #ara

Ready to run

npx skills add https://github.com/Aradotso/ara.engineer --skill braintrust

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: braintrust
Source: https://github.com/Aradotso/ara.engineer/tree/main/skills/braintrust
Command: npx skills add https://github.com/Aradotso/ara.engineer --skill braintrust

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Braintrust provides a CLI-based instrumentation and tracing toolkit to observe Ara services and their interactions across the stack.

Core Features & Use Cases

Cross-service tracing and log viewing across Ara components (website-agent, ara-api, ara.engineer) from a single project context.
SQL-like querying and drill-downs to identify bottlenecks, errors, and usage patterns.
Guidelines and templates for instrumenting new services and wrapping code with tracing.

Quick Start

Install Braintrust, authenticate with your Braintrust API key, and run a trace view to begin debugging Ara services.

Frequently Asked Questions about braintrust

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I debug issues across multiple Ara services at once?▼

Cross-service debugging is handled by tracing interactions across Ara components like website-agent and ara-api from a single project context. You can view logs and traces together to identify where errors or bottlenecks originate across the stack.

Can I query Ara service logs using SQL-like syntax?▼

SQL-like querying is supported to filter and drill down into Ara service traces and logs. You can write structured queries to identify specific errors, usage patterns, or performance bottlenecks across the traced components.

What's the best way to instrument new code for tracing within the Ara stack?▼

Structured tracing conventions and templates are provided to guide instrumenting new Ara services. These workflows ensure your code wrapping aligns with the existing Ara stack tracing standards for consistent observability.

Do I need a specific CLI setup to start inspecting Ara traces?▼

A CLI-based toolkit is used to inspect traces, requiring you to install the tool and authenticate with a Braintrust API key. Once configured, you can run trace views directly from your terminal to begin analyzing service interactions.

Why are my Ara component logs disconnected when I try to trace a request?▼

Disconnected logs often occur when tracing conventions are not enforced across services. The toolkit provides structured guidelines for wrapping code, ensuring that traces propagate correctly through website-agent, ara-api, and ara.engineer components.

Does this tracing toolkit work for analyzing performance bottlenecks in Ara services?▼

Performance bottleneck analysis is a core use case, allowing you to run drill-downs on service traces to pinpoint slow operations. By viewing cross-service interactions, you can isolate exactly which Ara component is causing the latency.