5 pro tips for using the Braintrust CLI

Braintrust’s bt command-line interface gives coding agents and developers terminal access to production logs, traces, datasets, evaluations, and deployed functions, enabling them to investigate real application behavior and validate proposed fixes. The workflow emphasizes running evals continuously in watch mode while editing, using stable datasets and scorers to isolate the effects of changes, and creating repeatable subset-based smoke checks for faster CI feedback while retaining full-suite tests before release. It also describes using log and SQL queries to identify slow requests and trace repeated tool calls, then converting production failures into reviewed regression-test dataset entries. Keeping scorers, tools, and LLM functions in the application repository supports version control, code review, and clearer interpretation of score changes, while CLI options enable automation in macOS or Linux CI environments through API-key authentication, noninteractive execution, JSON output, and custom reporters that can turn evaluation scores into pass/fail signals.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.