# Braintrust

> Braintrust is the active observability platform for instrumenting, understanding, and improving agents. By actively applying intelligence to agent traces and automatically surfacing the most critical patterns, Braintrust gives teams the visibility to understand how agents behave in production and the tools to improve them.

## Start

- [Get started with Braintrust](https://braintrust.dev/docs/index.md)
- [Braintrust workflow](https://braintrust.dev/docs/workflow.md): Understand how to trace, evaluate, and improve AI applications with Braintrust
- [Tracing quickstart](https://braintrust.dev/docs/tracing-quickstart.md): Instrument your AI application with a coding agent and verify that traces appear in Braintrust.
- [Evaluation quickstart](https://braintrust.dev/docs/evaluation-quickstart.md): Systematically measure AI quality. Compare approaches, catch regressions before deployment, and validate improvements with data instead of intuition.
- [Loop agent](https://braintrust.dev/docs/loop.md): Use natural language to analyze logs, optimize prompts, build datasets, and more
- [Plans and limits](https://braintrust.dev/docs/plans-and-limits.md): Choose the right Braintrust plan for your needs

## Instrument

- [Instrument your application](https://braintrust.dev/docs/instrument/index.md): Add tracing to your AI application to capture LLM calls, application logic, and user feedback — the foundation for observability and evaluation.

## Tracing

- [Trace LLM calls](https://braintrust.dev/docs/instrument/trace-llm-calls.md): Enable auto-instrumentation once and every call to a supported AI provider is logged — inputs, outputs, latency, token usage, and costs.
- [Trace application logic](https://braintrust.dev/docs/instrument/trace-application-logic.md): Trace retrieval steps, tool calls, preprocessing, and business logic in spans to get full end-to-end visibility into your AI application's behavior.
- [Advanced tracing patterns](https://braintrust.dev/docs/instrument/advanced-tracing.md): Implement distributed tracing across services, mask sensitive data, customize span rendering, and tune performance for high-throughput AI applications.

## Instrument

- [Capture user feedback](https://braintrust.dev/docs/instrument/user-feedback.md): Attach ratings, corrections, and comments from users to traces to measure real-world quality and power feedback-driven evaluations.
- [Log attachments](https://braintrust.dev/docs/instrument/attachments.md): Attach images, audio, PDFs, and other binary files to traces and experiments to review multimodal inputs and outputs in context.

## Observe

- [Observe your application](https://braintrust.dev/docs/observe/index.md): Turn every production request into a searchable trace. Filter, analyze, and drill into real-time data to identify issues and gather signal for improving your AI system.

## Logs

- [View your logs](https://braintrust.dev/docs/observe/view-logs.md): Browse and inspect production traces, create custom columns to surface relevant data, and organize logs with tags for faster triage and analysis.
- [Filter and search logs](https://braintrust.dev/docs/observe/filter.md): Narrow down to exactly the traces you need using metadata filters, SQL queries, or the API. Pull out specific production data for debugging, analysis, or dataset creation.
- [Examine traces](https://braintrust.dev/docs/observe/examine-traces.md): Open any trace to inspect its spans, inputs, outputs, and timing. Explore trace views, translate content, navigate origins, and share results.

## Topics

- [Topics](https://braintrust.dev/docs/observe/topics/index.md): Automatically classify logs by user intent, sentiment, and issues, then turn patterns into evaluation datasets, scorers, and review queues.
- [Enable Topics](https://braintrust.dev/docs/observe/topics/enable.md): Confirm your traces render with the Thread preprocessor, enable Topics, and adapt traces that need a custom preprocessor.
- [Review insights from Topics](https://braintrust.dev/docs/observe/topics/review-insights.md): Find patterns in classified logs through topic distributions, trace classifications, and time-series trends, with snapshot and real-time views.
- [Act on findings from Topics](https://braintrust.dev/docs/observe/topics/act-on-findings.md): Turn topic patterns into action: build evaluation datasets, set up online scoring by classification, and assign matching logs for human review.
- [Manage Topics](https://braintrust.dev/docs/observe/topics/manage.md): Keep the Topics pipeline healthy: check status, trigger runs on demand, adjust sampling, pause, and rewind history after facet or filter changes.
- [Custom facets](https://braintrust.dev/docs/observe/topics/custom-facets.md): Define domain-specific facets with custom preprocessors and prompts to extract patterns the built-in Task, Sentiment, and Issues facets miss.

## Observe

- [Monitor with dashboards](https://braintrust.dev/docs/observe/dashboards.md): Build custom dashboards to track request counts, latency, costs, scores, and other metrics across your logs and experiments over time.
- [Set up alerts](https://braintrust.dev/docs/observe/alerts.md): Configure webhooks and Slack notifications

## Annotate

- [Annotate and curate data](https://braintrust.dev/docs/annotate/index.md): Transform production logs into high-quality datasets. Add human feedback, labels, and corrections to build test cases that drive systematic improvement.

## Human review

- [Set up human review](https://braintrust.dev/docs/annotate/human-review/index.md): Capture structured human judgment on production traces to build ground truth, validate automated scores, and surface edge cases your scorers miss.
- [Score traces and datasets](https://braintrust.dev/docs/annotate/human-review/score-traces.md): Open logs, experiments, and dataset rows in review mode, set scores, add comments, and move through rows efficiently to record human judgments.
- [Review with multiple reviewers](https://braintrust.dev/docs/annotate/human-review/multiple-reviewers.md): See how Braintrust combines scores when several reviewers grade the same span, set a base value for non-numeric scores, and switch between reviewers.
- [Manage review work](https://braintrust.dev/docs/annotate/human-review/manage-review-work.md): Assign rows to reviewers, filter and organize the review queue, and customize table views to distribute and track human review across your team.

## Annotate

- [Create custom views](https://braintrust.dev/docs/annotate/custom-views.md): Help your team understand and act on traces and dataset rows faster. Describe the interface you need in natural language, and Loop generates a customizable React component you can embed anywhere.
- [Add labels and corrections](https://braintrust.dev/docs/annotate/labels.md): Go beyond numeric scores by tagging traces, adding expected values, and writing corrections. Use these annotations to organize data and build higher-quality datasets.

## Datasets

- [Datasets](https://braintrust.dev/docs/annotate/datasets/index.md): Versioned collections of test cases that power repeatable evaluations and capture real production behavior as your application evolves.
- [Create datasets](https://braintrust.dev/docs/annotate/datasets/create.md): Build datasets from CSV uploads, the SDK, production logs, user feedback, traces, or Loop. Includes multimodal records like images and attachments.
- [Dataset pipelines](https://braintrust.dev/docs/annotate/datasets/pipelines.md): Transform spans or traces from your project logs into dataset rows in bulk with declarative, version-controlled pipelines.
- [Manage datasets](https://braintrust.dev/docs/annotate/datasets/manage.md): Organize datasets with tags and stars, pin versions with snapshots, validate with schemas, and edit records programmatically.
- [Use datasets in evaluations](https://braintrust.dev/docs/annotate/datasets/use-in-evaluations.md): Pass datasets directly to Eval(), assign them to environments, or convert experiment results into dataset format.
- [Track dataset performance](https://braintrust.dev/docs/annotate/datasets/track-performance.md): See which experiments used a dataset and how each row performs over time, with chart and SQL filters to spot trends.

## Annotate

- [Export annotated data](https://braintrust.dev/docs/annotate/export.md): Take annotated traces and datasets outside Braintrust to use in external evaluation frameworks, analysis pipelines, reporting, or training workflows.

## Evaluate

- [Evaluate systematically](https://braintrust.dev/docs/evaluate/index.md): Measure AI application quality, detect regressions before they reach production, and build confidence that your system is improving over time.
- [Create prompts](https://braintrust.dev/docs/evaluate/write-prompts.md): Version and deploy prompts independently of your code. Test changes in the playground and roll out to specific environments without redeploying your application.

## Scorers and classifiers

- [Evaluate output quality with scorers and classifiers](https://braintrust.dev/docs/evaluate/write-scorers.md): Build scorers and classifiers to evaluate AI outputs with numeric scores or categorical labels, in experiments and in production.
- [Autoevals](https://braintrust.dev/docs/evaluate/autoevals.md): Use pre-built, battle-tested scorers for common evaluation tasks like factuality checking, semantic similarity, and format validation.
- [Custom code scorers and classifiers](https://braintrust.dev/docs/evaluate/custom-code.md): Write scorers and classifiers as custom code with full control over scoring logic, business rules, and pattern matching to evaluate AI outputs.
- [LLM-as-a-judge scorers and classifiers](https://braintrust.dev/docs/evaluate/llm-as-a-judge.md): Use language models to score or classify AI outputs based on natural language criteria like tone, helpfulness, or intent.

## Evaluate

- [Iterate in playgrounds](https://braintrust.dev/docs/evaluate/playgrounds.md): Prototype prompts, swap models, and test scorers interactively. Validate ideas quickly in the browser before committing to a full evaluation experiment.

## Experiments

- [Experiments](https://braintrust.dev/docs/evaluate/run-evaluations.md): Choose how to run evaluations, in code, the UI, or CI/CD, and turn each run into a permanent record you can compare and share.
- [Run experiments in code](https://braintrust.dev/docs/evaluate/run-in-code.md): Run evaluations with the Braintrust SDK, from a local script or the bt eval CLI, and capture each run as an immutable experiment snapshot.
- [Run experiments in the UI](https://braintrust.dev/docs/evaluate/run-in-ui.md): Evaluate prompts and workflows in the Braintrust UI, with no code or local setup required.
- [Run experiments in CI/CD](https://braintrust.dev/docs/evaluate/run-in-ci.md): Automate evaluation runs in CI with GitHub Actions or the bt eval CLI to keep regressions off your main branch.
- [Advanced eval techniques](https://braintrust.dev/docs/evaluate/advanced-evaluations.md): Improve scoring reliability with trials and hill climbing, evaluate images and PDFs with attachments, and trace tasks to debug them.

## Analyze results

- [Interpret evaluation results](https://braintrust.dev/docs/evaluate/interpret-results.md): Diagnose where your AI system is underperforming and understand why. Drill into traces, score distributions, and individual test cases to find actionable improvement opportunities.
- [Compare experiments](https://braintrust.dev/docs/evaluate/compare-experiments.md): Identify which test cases improved or regressed between runs. Set a baseline experiment to align test cases and compare score deltas side by side.
- [Evaluate reasoning models](https://braintrust.dev/docs/evaluate/reasoning.md): Configure thinking budgets, capture intermediate reasoning traces, and score final outputs for models like OpenAI o4, Claude, and Gemini that think before responding.

## Remote evals

- [Remote evals and sandboxes](https://braintrust.dev/docs/evaluate/remote-evals.md): Connect custom agent code to a playground, run it in a sandbox, and test complex agents interactively without modifying your deployed code.
- [Create evaluation parameters](https://braintrust.dev/docs/evaluate/write-parameters.md): Store configurable settings as versioned parameters and reference them across experiments, prompts, and environments without hardcoding values in your code.

## Evaluate

- [Score production traces](https://braintrust.dev/docs/evaluate/score-online.md): Get continuous quality monitoring on every production request. Evaluations run asynchronously in the background without adding latency to your application.
- [Evaluation best practices](https://braintrust.dev/docs/evaluate/best-practices.md): Build evaluations that produce reliable, actionable signals. Learn how to structure datasets, write effective scorers, and avoid common pitfalls.

## Deploy

- [Deploy](https://braintrust.dev/docs/deploy/index.md): Ship your AI applications to production with confidence
- [Deploy prompts](https://braintrust.dev/docs/deploy/prompts.md): Ship and version prompts in production
- [Deploy functions](https://braintrust.dev/docs/deploy/functions.md): Deploy tools, scorers, and workflows to production

## Braintrust Gateway

- [Use the Braintrust Gateway](https://braintrust.dev/docs/deploy/gateway.md): Route requests to any AI provider through a unified LLM API
- [Supported models](https://braintrust.dev/docs/deploy/supported-models.md): Models available through the Braintrust gateway
- [Use the proxy (deprecated)](https://braintrust.dev/docs/deploy/ai-proxy.md): Call any AI provider through a unified interface (deprecated, use the gateway instead)

## Deploy

- [Manage environments](https://braintrust.dev/docs/deploy/environments.md): Separate dev, staging, and production configurations
- [Monitor deployments](https://braintrust.dev/docs/deploy/monitor.md): Track production performance and errors
- [Stream responses](https://braintrust.dev/docs/deploy/streaming.md): Return incremental responses from prompts and functions

## Admin

- [Administration](https://braintrust.dev/docs/admin/index.md): Manage organizations, access control, and infrastructure
- [Manage organizations](https://braintrust.dev/docs/admin/organizations.md): Configure organization settings and integrations
- [Manage projects](https://braintrust.dev/docs/admin/projects.md): Configure project settings and features
- [Configure AI providers](https://braintrust.dev/docs/admin/ai-providers.md): Manage organization-level and project-level AI provider credentials for playgrounds, experiments, and the gateway.
- [Audit logging](https://braintrust.dev/docs/admin/audit-logs.md): Track and review administrative actions in your Braintrust organization, such as permission changes, member management, and API key creation

## Access control

- [Access control](https://braintrust.dev/docs/admin/access-control/index.md): Control who can access your projects, experiments, and data using permission groups scoped to the organization, project, and object levels
- [Manage permissions](https://braintrust.dev/docs/admin/access-control/manage-permissions.md): Create permission groups, assign permissions, and provision service accounts

## Billing

- [Manage billing](https://braintrust.dev/docs/admin/billing/index.md): Change plans, monitor usage, and manage billing
- [Change your plan](https://braintrust.dev/docs/admin/billing/change-plan.md): Upgrade or downgrade between Starter, Pro, and Enterprise
- [Monitor usage](https://braintrust.dev/docs/admin/billing/monitor-usage.md): Track usage and set up cost alerts
- [Billing FAQ](https://braintrust.dev/docs/admin/billing/faq.md): Common questions about Braintrust pricing, billing, and plans

## Data management

- [Export to cloud storage](https://braintrust.dev/docs/admin/data-management/export.md): Deliver project logs to S3 or GCS in JSONL or Parquet format, with Hive partitioning for efficient downstream querying.
- [Configure data retention](https://braintrust.dev/docs/admin/data-management/retention.md): Extend your retention window to keep data longer, or automate and manually delete data to meet compliance obligations and store less.

## Deployment

- [Deployment options](https://braintrust.dev/docs/admin/deployment/index.md): Compare SaaS, BYOC, and self-hosted deployments and choose the right data residency and operational model for your team.
- [Bring your own cloud (BYOC)](https://braintrust.dev/docs/admin/deployment/byoc.md): Have Braintrust operate the data plane inside your own cloud account, keeping data in your boundary without owning day-2 operations.

## Self-hosting

- [Self-hosting Braintrust](https://braintrust.dev/docs/admin/self-hosting/index.md)
- [Architecture](https://braintrust.dev/docs/admin/self-hosting/architecture.md)
- [Deploy Braintrust](https://braintrust.dev/docs/admin/self-hosting/deploy.md): Deploy a self-hosted Braintrust data plane on AWS, GCP, or Azure

## Configure

- [Configure your deployment](https://braintrust.dev/docs/admin/self-hosting/configure/index.md): Tune a self-hosted Braintrust data plane after deployment with security, networking, scaling, and telemetry settings.
- [Security and access](https://braintrust.dev/docs/admin/self-hosting/configure/security.md): Restrict which organizations can authenticate, secure customer data, validate outbound URLs, audit data access, and harden pods.
- [Networking and connectivity](https://braintrust.dev/docs/admin/self-hosting/configure/networking.md): Configure URLs, rate limits, load balancer settings, network access paths, and the Gateway for a self-hosted Braintrust data plane.
- [Scaling and storage](https://braintrust.dev/docs/admin/self-hosting/configure/scaling.md): Size the API and Brainstore compute in a self-hosted data plane with Terraform variables and Helm resource settings.
- [Telemetry and data retention](https://braintrust.dev/docs/admin/self-hosting/configure/telemetry.md): Control the telemetry a self-hosted data plane sends to Braintrust's control plane and enable data retention.

## Upgrade

- [Upgrade your deployment](https://braintrust.dev/docs/admin/self-hosting/upgrade/routine.md): Upgrade a self-hosted Braintrust data plane on AWS, GCP, or Azure

## Version migrations

- [Upgrade to data plane v2.x](https://braintrust.dev/docs/admin/self-hosting/upgrade/v2.md): Upgrade your self-hosted Braintrust deployment to the data plane v2 series
- [Upgrade to Terraform module v6](https://braintrust.dev/docs/admin/self-hosting/upgrade/v6.md): Upgrade your self-hosted Braintrust AWS deployment from Terraform module v5.x to v6 and cut API traffic over from Lambda to ECS.

## Admin

- [Authentication](https://braintrust.dev/docs/admin/authentication.md): Understand how users and services authenticate to Braintrust, and how Braintrust authenticates to model providers on your behalf.
- [Personal settings](https://braintrust.dev/docs/admin/personal-settings.md): Configure your individual user preferences
- [Service health](https://braintrust.dev/docs/admin/service-health.md): Monitor Braintrust service availability and subscribe to incident notifications

## Best practices

- [Evaluating agents](https://braintrust.dev/docs/best-practices/agents.md)
- [Writing scorers](https://braintrust.dev/docs/best-practices/scorers.md)
- [Product manager workflows](https://braintrust.dev/docs/best-practices/pm-workflows.md)

## Guides

- [Security](https://braintrust.dev/docs/security.md)

## Integrations

- [Integrations](https://braintrust.dev/docs/integrations/index.md): Connect Braintrust with your AI providers and frameworks

## AI providers

- [AI providers](https://braintrust.dev/docs/integrations/ai-providers/index.md): AI provider configuration

## Model providers

- [OpenAI](https://braintrust.dev/docs/integrations/ai-providers/openai.md): Trace OpenAI SDK calls, evaluate GPT models, and route them through the Braintrust gateway
- [OpenRouter](https://braintrust.dev/docs/integrations/ai-providers/openrouter.md): Trace OpenRouter SDK calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Anthropic](https://braintrust.dev/docs/integrations/ai-providers/anthropic.md): Trace Anthropic SDK calls, evaluate Claude models, and route them through the Braintrust gateway
- [Gemini](https://braintrust.dev/docs/integrations/ai-providers/gemini.md): Trace Gemini SDK calls, run evals, and use Gemini models in Braintrust
- [Mistral](https://braintrust.dev/docs/integrations/ai-providers/mistral.md): Trace Mistral SDK calls and evaluate Mistral models in Braintrust
- [Baseten](https://braintrust.dev/docs/integrations/ai-providers/baseten.md): Baseten model provider configuration and integration guide
- [Cerebras](https://braintrust.dev/docs/integrations/ai-providers/cerebras.md): Cerebras model provider configuration and integration guide
- [Cohere](https://braintrust.dev/docs/integrations/ai-providers/cohere.md): Trace Cohere SDK calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Fireworks](https://braintrust.dev/docs/integrations/ai-providers/fireworks.md): Fireworks AI model provider configuration and integration guide
- [Groq](https://braintrust.dev/docs/integrations/ai-providers/groq.md): Trace Groq SDK calls and use Groq models in Braintrust
- [Hugging Face](https://braintrust.dev/docs/integrations/ai-providers/huggingface.md): Trace Hugging Face Inference, Transformers.js, and local Transformers pipeline calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Lepton](https://braintrust.dev/docs/integrations/ai-providers/lepton.md): Lepton AI model provider configuration and integration guide
- [Ollama](https://braintrust.dev/docs/integrations/ai-providers/ollama.md): Trace Ollama SDK calls in Braintrust to debug local model inference for chat, generate, and embedding calls
- [Perplexity](https://braintrust.dev/docs/integrations/ai-providers/perplexity.md): Perplexity AI model provider configuration and integration guide
- [Replicate](https://braintrust.dev/docs/integrations/ai-providers/replicate.md): Replicate model provider configuration and integration guide
- [Together](https://braintrust.dev/docs/integrations/ai-providers/together.md): Together AI model provider configuration and integration guide
- [Voyage AI](https://braintrust.dev/docs/integrations/ai-providers/voyageai.md): Trace Voyage AI SDK calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [xAI](https://braintrust.dev/docs/integrations/ai-providers/xai.md): xAI model provider configuration and integration guide

## Cloud providers

- [AWS Bedrock](https://braintrust.dev/docs/integrations/ai-providers/bedrock.md): Trace AWS Bedrock SDK calls and use Bedrock models in Braintrust
- [Google Vertex AI](https://braintrust.dev/docs/integrations/ai-providers/google.md): Configure Google Cloud Vertex AI in Braintrust to access Google and partner models
- [Azure AI Foundry](https://braintrust.dev/docs/integrations/ai-providers/azure.md): Trace Azure OpenAI SDK calls and use Azure AI Foundry models in Braintrust
- [Azure AI Gateway](https://braintrust.dev/docs/integrations/ai-providers/azure-ai-gateway.md): Use the models behind your Azure AI Gateway in Braintrust playgrounds, prompts, and scorers
- [Databricks](https://braintrust.dev/docs/integrations/ai-providers/databricks.md): Configure Databricks Model Serving to access foundation models

## AI providers

- [Custom providers](https://braintrust.dev/docs/integrations/ai-providers/custom.md): Set up custom AI providers with Braintrust for evaluation and tracing

## SDKs

- [SDK integrations](https://braintrust.dev/docs/integrations/sdk-integrations/index.md): Integrate with existing frameworks

## Tracing integrations

- [OpenTelemetry (OTel)](https://braintrust.dev/docs/integrations/sdk-integrations/opentelemetry.md)
- [Vercel](https://braintrust.dev/docs/integrations/sdk-integrations/vercel.md): Trace Vercel AI SDK calls and collect Vercel Marketplace traces in Braintrust
- [LangChain](https://braintrust.dev/docs/integrations/sdk-integrations/langchain.md): Trace LangChain applications in Braintrust to debug prompts, evaluate models, and monitor production usage
- [LangSmith](https://braintrust.dev/docs/integrations/sdk-integrations/langsmith.md): Trace LangSmith applications in Braintrust to observe and evaluate LLM calls, chains, and tools
- [LlamaIndex](https://braintrust.dev/docs/integrations/sdk-integrations/llamaindex.md): Trace LlamaIndex calls in Braintrust to debug prompts, evaluate models, and monitor RAG pipelines
- [Agno](https://braintrust.dev/docs/integrations/sdk-integrations/agno.md): Trace Agno agents, teams, and workflows in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Apollo GraphQL](https://braintrust.dev/docs/integrations/sdk-integrations/apollo-graphql.md)
- [Cloudflare](https://braintrust.dev/docs/integrations/sdk-integrations/cloudflare.md): Trace AI requests through Cloudflare AI Gateway and export Cloudflare Agents traces to Braintrust with OpenTelemetry
- [CloudWeGo Eino](https://braintrust.dev/docs/integrations/sdk-integrations/cloudwego-eino.md): Trace CloudWeGo Eino chat model, tool, and embedding calls in Braintrust to debug and evaluate your LLM apps
- [DSPy](https://braintrust.dev/docs/integrations/sdk-integrations/dspy.md): Trace DSPy pipelines in Braintrust to debug modules, evaluate prompts, and monitor LLM calls
- [Firebase Genkit](https://braintrust.dev/docs/integrations/sdk-integrations/firebase-genkit.md): Trace Firebase Genkit calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Instructor](https://braintrust.dev/docs/integrations/sdk-integrations/instructor.md): Trace Instructor structured-output calls in Braintrust to debug extractions, evaluate models, and monitor retries
- [LiteLLM](https://braintrust.dev/docs/integrations/sdk-integrations/litellm.md): Trace LiteLLM calls in Braintrust to debug routing and outputs across 100+ model providers
- [RubyLLM](https://braintrust.dev/docs/integrations/sdk-integrations/ruby-llm.md): Trace RubyLLM applications in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Spring AI](https://braintrust.dev/docs/integrations/sdk-integrations/spring-ai.md): Trace Spring AI calls in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Temporal](https://braintrust.dev/docs/integrations/sdk-integrations/temporal.md): Trace Temporal workflow and activity executions in Braintrust to debug and monitor distributed applications
- [TraceLoop](https://braintrust.dev/docs/integrations/sdk-integrations/traceloop.md)
- [TrueFoundry](https://braintrust.dev/docs/integrations/sdk-integrations/truefoundry.md): Export LLM traces from TrueFoundry AI Gateway to Braintrust

## Agent frameworks

- [OpenAI Agents SDK](https://braintrust.dev/docs/integrations/agent-frameworks/openai-agents-sdk.md): Trace OpenAI Agents SDK runs in Braintrust to debug agents, evaluate models, and monitor production
- [OpenAI Codex SDK](https://braintrust.dev/docs/integrations/agent-frameworks/openai-codex-sdk.md): Trace OpenAI Codex SDK runs in Braintrust to debug agent turns, tool calls, and code-editing workflows
- [Pi Coding Agent SDK](https://braintrust.dev/docs/integrations/agent-frameworks/pi-coding-agent.md): Trace Pi Coding Agent SDK prompts, Anthropic LLM calls, and tool invocations in Braintrust
- [OpenRouter Agent](https://braintrust.dev/docs/integrations/agent-frameworks/openrouter-agent.md): Trace OpenRouter Agent calls with Braintrust
- [Claude Agent SDK](https://braintrust.dev/docs/integrations/agent-frameworks/claude-agent-sdk.md): Trace Claude Agent SDK runs in Braintrust to debug agent queries, tool calls, and multi-step reasoning
- [Cursor SDK](https://braintrust.dev/docs/integrations/agent-frameworks/cursor-sdk.md): Trace Cursor agent SDK runs in Braintrust to debug tool calls, subagents, and multi-step reasoning
- [Cloudflare Think](https://braintrust.dev/docs/integrations/agent-frameworks/cloudflare-think.md): Trace @cloudflare/think agent runs in Braintrust to debug multi-turn reasoning, LLM calls, and tool use on Cloudflare Workers
- [Flue](https://braintrust.dev/docs/integrations/agent-frameworks/flue.md): Trace Flue agent workflows in Braintrust to debug LLM turns, tool calls, delegated tasks, and context compaction
- [Deep Agents](https://braintrust.dev/docs/integrations/agent-frameworks/deepagents.md): Trace Deep Agents workflows in Braintrust to debug agent plans, tool calls, and model calls. Uses the Braintrust LangChain integration with auto-instrumentation or manual callbacks.
- [Eve](https://braintrust.dev/docs/integrations/agent-frameworks/eve.md): Trace Eve agent sessions in Braintrust to observe LLM steps, tool calls, and subagent interactions
- [LangGraph](https://braintrust.dev/docs/integrations/agent-frameworks/langgraph.md)
- [CrewAI](https://braintrust.dev/docs/integrations/agent-frameworks/crew-ai.md): Trace CrewAI multi-agent runs in Braintrust to debug crews, agents, and tool calls
- [AutoGen](https://braintrust.dev/docs/integrations/agent-frameworks/autogen.md): Trace AutoGen multi-agent runs in Braintrust to debug agents, teams, and tool calls
- [AgentScope](https://braintrust.dev/docs/integrations/agent-frameworks/agentscope.md): Trace AgentScope multi-agent runs in Braintrust to debug agents, pipelines, and tool calls
- [Google ADK (Agent Development Kit)](https://braintrust.dev/docs/integrations/agent-frameworks/google.md): Trace Google ADK agents in Braintrust to debug prompts, tool calls, and multi-step reasoning
- [LiveKit Agents](https://braintrust.dev/docs/integrations/agent-frameworks/livekit-agents.md): Trace real-time voice AI pipelines built with LiveKit Agents, including LLM turns, speech-to-text, text-to-speech, and function tool calls
- [Pipecat](https://braintrust.dev/docs/integrations/agent-frameworks/pipecat.md): Trace Pipecat AI voice pipelines in Braintrust to debug LLM turns, speech-to-text, text-to-speech, and tool calls
- [Mastra](https://braintrust.dev/docs/integrations/agent-frameworks/mastra.md)
- [Microsoft Agent Framework](https://braintrust.dev/docs/integrations/agent-frameworks/microsoft-agent-framework.md): Trace Microsoft Agent Framework agents in Braintrust to debug agent runs, model calls, and tool executions
- [Pydantic AI](https://braintrust.dev/docs/integrations/agent-frameworks/pydantic-ai.md): Trace Pydantic AI agents in Braintrust to debug prompts, evaluate models, and monitor production usage
- [Strands Agents SDK](https://braintrust.dev/docs/integrations/agent-frameworks/strands-agent.md): Trace Strands agent runs in Braintrust to debug agent reasoning, model calls, and tool execution
- [Cloudflare Agents](https://braintrust.dev/docs/integrations/agent-frameworks/cloudflare-agents.md): Trace Cloudflare Agents tool calls in Braintrust to debug agent behavior, inspect tool inputs and outputs, and monitor errors
- [Cloudflare AI Chat](https://braintrust.dev/docs/integrations/agent-frameworks/cloudflare-ai-chat.md): Trace Cloudflare AI Chat agent conversations in Braintrust to debug chat turns, LLM calls, tool use, and errors

## Eval frameworks

- [Harbor](https://braintrust.dev/docs/integrations/sdk-integrations/harbor.md): Sync Harbor agent evaluation results to Braintrust with the native Harbor job plugin, capturing datasets, experiments, scores, and agent traces

## Testing frameworks

- [Pytest](https://braintrust.dev/docs/integrations/sdk-integrations/pytest.md)
- [Node.js test runner](https://braintrust.dev/docs/integrations/sdk-integrations/node-test-runner.md)
- [Vitest](https://braintrust.dev/docs/integrations/sdk-integrations/vitest.md): Run Braintrust evaluations from Vitest and report test results as experiments

## Developer tools

- [Braintrust MCP](https://braintrust.dev/docs/integrations/developer-tools/mcp.md): Connect Claude Code, Cursor, Codex, and other MCP clients to Braintrust. Query logs, author scorers, configure Topics, and run evals from your editor.
- [Claude Code](https://braintrust.dev/docs/integrations/developer-tools/claude-code.md): Connect Braintrust to Claude Code in the terminal or desktop app with the braintrust and trace-claude-code plugins to query data and trace sessions.
- [Claude Cowork](https://braintrust.dev/docs/integrations/developer-tools/claude-cowork.md): Connect Braintrust to Claude Cowork with the braintrust and trace-claude-code plugins to query your data and trace sessions.
- [Cursor](https://braintrust.dev/docs/integrations/developer-tools/cursor.md)
- [VS Code](https://braintrust.dev/docs/integrations/developer-tools/vscode.md)
- [Devin Desktop](https://braintrust.dev/docs/integrations/developer-tools/devin-desktop.md): Connect the Braintrust MCP server to Devin Desktop so its coding agent can access your projects, experiments, and logs.
- [Codex](https://braintrust.dev/docs/integrations/developer-tools/codex.md): Trace Codex CLI sessions to Braintrust and access your Braintrust data in Codex through the MCP server
- [OpenCode](https://braintrust.dev/docs/integrations/developer-tools/opencode.md)
- [pi](https://braintrust.dev/docs/integrations/developer-tools/pi.md)

## Runtime environments

- [AWS Lambda](https://braintrust.dev/docs/integrations/runtime-environments/aws-lambda.md): Reduce request-path logging latency from AWS Lambda functions with the Braintrust Lambda Extension

## All SDKs

- [SDKs](https://braintrust.dev/docs/sdks/index.md): Build, evaluate, and observe AI applications with Braintrust in your own code.

## TypeScript

- [TypeScript SDK quickstart](https://braintrust.dev/docs/sdks/typescript/quickstart.md): Install the Braintrust TypeScript SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/typescript/install-and-instrument.md): Install the TypeScript SDK, set an API key, and configure tracing by hand.
- [TypeScript SDK integrations](https://braintrust.dev/docs/sdks/typescript/sdk-integrations.md): AI providers, agent frameworks, and tools the Braintrust TypeScript SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/typescript/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust TypeScript SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/typescript/api-reference.md): Functions, classes, and configuration options found in the Braintrust TypeScript SDK.

## Versions

- [Braintrust TypeScript SDK](https://braintrust.dev/docs/sdks/typescript/versions/3.27.0.md): TypeScript API reference for Braintrust SDK v3.27.0

## Migration guides

- [Migrate from v2.x to v3.x](https://braintrust.dev/docs/sdks/typescript/migrations/v2-to-v3.md): Guide for migrating from TypeScript SDK v2.x to v3.x
- [Migrate from v1.x to v2.x](https://braintrust.dev/docs/sdks/typescript/migrations/v1-to-v2.md): Guide for migrating from TypeScript SDK v1.x to v2.x
- [Migrate from v0.x to v1.x](https://braintrust.dev/docs/sdks/typescript/migrations/v0-to-v1.md): Guide for migrating from TypeScript SDK v0.x to v1.x

## Reference

- [Changelog](https://braintrust.dev/docs/sdks/typescript/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/typescript/github.md)

## Related packages

- [TypeScript SDK: related packages](https://braintrust.dev/docs/sdks/typescript/related/index.md): Adjacent npm packages that work alongside the Braintrust TypeScript SDK.

## Autoevals

- [Autoevals TypeScript API](https://braintrust.dev/docs/sdks/typescript/related/autoevals/0.3.0.md): TypeScript API reference for Autoevals v0.3.0

## LangChain

- [LangChain JS Integration](https://braintrust.dev/docs/sdks/typescript/related/langchain/0.2.4.md): LangChain JS Integration v0.2.4

## OpenAI Agents

- [OpenAI Agents JS Integration](https://braintrust.dev/docs/sdks/typescript/related/openai-agents/0.1.5.md): OpenAI Agents JS Integration v0.1.5

## OpenTelemetry

- [OpenTelemetry JS Integration](https://braintrust.dev/docs/sdks/typescript/related/otel/0.3.0.md): OpenTelemetry JS Integration v0.3.0

## Python

- [Python SDK quickstart](https://braintrust.dev/docs/sdks/python/quickstart.md): Install the Braintrust Python SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/python/install-and-instrument.md): Install the Python SDK, set an API key, and configure tracing by hand.
- [Python SDK integrations](https://braintrust.dev/docs/sdks/python/sdk-integrations.md): AI providers, agent frameworks, and tools the Braintrust Python SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/python/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust Python SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/python/api-reference.md): Functions, classes, and configuration options found in the Braintrust Python SDK.

## Versions

- [Braintrust Python SDK](https://braintrust.dev/docs/sdks/python/versions/0.33.0.md): Python API reference for Braintrust SDK v0.33.0

## Reference

- [Changelog](https://braintrust.dev/docs/sdks/python/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/python/github.md)

## Related packages

- [Python SDK: related packages](https://braintrust.dev/docs/sdks/python/related/index.md): Adjacent Python packages that work alongside the Braintrust Python SDK.

## Autoevals

- [Autoevals Python API](https://braintrust.dev/docs/sdks/python/related/autoevals/0.3.0.md): Python API reference for Autoevals v0.3.0

## LangChain (deprecated)

- [LangChain Python Integration](https://braintrust.dev/docs/sdks/python/related/langchain/0.3.0.md): Braintrust integration for LangChain Python v0.3.0 (deprecated)

## Google ADK (deprecated)

- [Google ADK Integration](https://braintrust.dev/docs/sdks/python/related/adk/0.4.0.md): Braintrust integration for Google ADK v0.4.0 (deprecated)

## Go

- [Go SDK quickstart](https://braintrust.dev/docs/sdks/go/quickstart.md): Install the Braintrust Go SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/go/install-and-instrument.md): Install the Go SDK, set an API key, and configure tracing by hand.
- [Go SDK integrations](https://braintrust.dev/docs/sdks/go/sdk-integrations.md): AI providers and tools the Braintrust Go SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/go/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust Go SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/go/api-reference.md): Functions and configuration options found in the Braintrust Go SDK.

## Versions

- [Braintrust Go SDK](https://braintrust.dev/docs/sdks/go/versions/0.11.0.md): Go API reference for Braintrust SDK v0.11.0
- [Braintrust Go SDK](https://braintrust.dev/docs/sdks/go/versions/0.9.1.md): Go API reference for Braintrust SDK v0.9.1

## Migration guides

- [Migrate from v0.3.x to v0.4.x](https://braintrust.dev/docs/sdks/go/migrations/v0-3-to-v0-4.md): Guide for migrating from Go SDK v0.3.x to v0.4.x

## Reference

- [Changelog](https://braintrust.dev/docs/sdks/go/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/go/github.md)

## Java

- [Java SDK quickstart](https://braintrust.dev/docs/sdks/java/quickstart.md): Install the Braintrust Java SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/java/install-and-instrument.md): Install the Java SDK, set an API key, and configure tracing by hand.
- [Java SDK integrations](https://braintrust.dev/docs/sdks/java/sdk-integrations.md): AI providers and frameworks the Braintrust Java SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/java/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust Java SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/java/api-reference.md): Functions, classes, and configuration options found in the Braintrust Java SDK.

## Versions

- [0.3.21](https://braintrust.dev/docs/sdks/java/versions/0.3.21.md)

## Reference

- [Changelog](https://braintrust.dev/docs/sdks/java/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/java/github.md)

## Ruby

- [Ruby SDK quickstart](https://braintrust.dev/docs/sdks/ruby/quickstart.md): Install the Braintrust Ruby SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/ruby/install-and-instrument.md): Install the Ruby SDK, set an API key, and configure tracing by hand.
- [Ruby SDK integrations](https://braintrust.dev/docs/sdks/ruby/sdk-integrations.md): AI providers and tools the Braintrust Ruby SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/ruby/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust Ruby SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/ruby/api-reference.md): Functions, classes, and configuration options found in the Braintrust Ruby SDK.

## Versions

- [0.4.0](https://braintrust.dev/docs/sdks/ruby/versions/0.4.0.md)

## Reference

- [Changelog](https://braintrust.dev/docs/sdks/ruby/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/ruby/github.md)

## C#

- [C# SDK quickstart](https://braintrust.dev/docs/sdks/csharp/quickstart.md): Install the Braintrust .NET SDK, instrument your application, and verify that traces appear in Braintrust.

## Use the SDK

- [Install and instrument](https://braintrust.dev/docs/sdks/csharp/install-and-instrument.md): Install the .NET SDK, set an API key, and configure tracing by hand.
- [C# SDK integrations](https://braintrust.dev/docs/sdks/csharp/sdk-integrations.md): AI providers and tools the Braintrust .NET SDK integrates with.
- [Troubleshooting](https://braintrust.dev/docs/sdks/csharp/troubleshooting.md): Diagnose and fix common issues when installing or using the Braintrust .NET SDK.

## Reference

- [API reference](https://braintrust.dev/docs/sdks/csharp/api-reference.md): Functions, classes, and configuration options found in the Braintrust .NET SDK.
- [Changelog](https://braintrust.dev/docs/sdks/csharp/changelog.md)
- [GitHub](https://braintrust.dev/docs/sdks/csharp/github.md)

- [Reference (254 pages)](https://braintrust.dev/docs/_llms/reference.md): Documentation for Reference.

## Cookbook

- [Cookbook](https://braintrust.dev/docs/cookbook/index.md)

## Recipes

- [Supervising agent behavior with behavior specs](https://braintrust.dev/docs/cookbook/recipes/AgentBehavior.md)
- [Analyzing agent traces from Hugging Face](https://braintrust.dev/docs/cookbook/recipes/HuggingFaceAgentTraces.md)
- [Finding production issues using Topics](https://braintrust.dev/docs/cookbook/recipes/FindingProductionIssues.md)
- [Building a deep research agent with Temporal](https://braintrust.dev/docs/cookbook/recipes/TemporalDeepResearch.md)
- [Evaluating and iterating on AI apps with Lovable](https://braintrust.dev/docs/cookbook/recipes/Lovable.md)
- [Using Loop in AI product development](https://braintrust.dev/docs/cookbook/recipes/Loop.md)
- [Observability for Strands Agents on Amazon Bedrock](https://braintrust.dev/docs/cookbook/recipes/AmazonBedrockStrands.md)
- [Building reliable AI agents](https://braintrust.dev/docs/cookbook/recipes/AgentWhileLoop.md)
- [Using PDF attachments in playgrounds](https://braintrust.dev/docs/cookbook/recipes/PDFPlayground.md)
- [Tracing Vercel AI SDK applications](https://braintrust.dev/docs/cookbook/recipes/VercelAISDKTracing.md)
- [Evaluating video QA with Twelve Labs](https://braintrust.dev/docs/cookbook/recipes/VideoQATwelveLabs.md)
- [Evaluating a web agent](https://braintrust.dev/docs/cookbook/recipes/WebAgent.md)
- [Prompt versioning and deployment](https://braintrust.dev/docs/cookbook/recipes/PromptVersioning.md)
- [Evaluating video QA](https://braintrust.dev/docs/cookbook/recipes/VideoQA.md)
- [Evaluating a voice agent](https://braintrust.dev/docs/cookbook/recipes/VoiceAgent.md)
- [Classifying spam using structured outputs](https://braintrust.dev/docs/cookbook/recipes/SpamClassifier.md)
- [Evaluating a prompt chaining agent](https://braintrust.dev/docs/cookbook/recipes/PromptChaining.md)
- [Evaluating the precision and recall of an emotion classifier](https://braintrust.dev/docs/cookbook/recipes/PrecisionRecall.md)
- [Evaluating audio with the OpenAI Realtime API](https://braintrust.dev/docs/cookbook/recipes/Realtime.md)
- [Evaluating SimpleQA](https://braintrust.dev/docs/cookbook/recipes/SimpleQA.md)
- [Evaluating voice AI agents with Evalion](https://braintrust.dev/docs/cookbook/recipes/EvalionVoiceAgentEval.md)
- [Using Python functions to extract text from images](https://braintrust.dev/docs/cookbook/recipes/ToolOCR.md)
- [Using OpenTelemetry for LLM observability](https://braintrust.dev/docs/cookbook/recipes/OTEL-logging.md)
- [Using functions to build a RAG agent](https://braintrust.dev/docs/cookbook/recipes/ToolRAG.md)
- [Evaluating multimodal receipt extraction](https://braintrust.dev/docs/cookbook/recipes/ReceiptExtraction.md)
- [Unreleased AI: A full stack Next.js app for generating changelogs](https://braintrust.dev/docs/cookbook/recipes/UnreleasedAI.md)
- [An agent that runs OpenAPI commands](https://braintrust.dev/docs/cookbook/recipes/APIAgent-Py.md)
- [Benchmarking inference providers](https://braintrust.dev/docs/cookbook/recipes/ProviderBenchmark.md)
- [Tool calls in LLaMa 3.1](https://braintrust.dev/docs/cookbook/recipes/LLaMa-3_1-Tools.md)
- [Evaluating a chat assistant](https://braintrust.dev/docs/cookbook/recipes/EvaluatingChatAssistant.md)
- [LLM Eval For Text2SQL](https://braintrust.dev/docs/cookbook/recipes/Text2SQL-Data.md)
- [Optimizing Ragas to evaluate a RAG pipeline](https://braintrust.dev/docs/cookbook/recipes/SimpleRagas.md)
- [Comparing evals across multiple AI models](https://braintrust.dev/docs/cookbook/recipes/ModelComparison.md)
- [Detecting Prompt Injections](https://braintrust.dev/docs/cookbook/recipes/PromptInjectionDetector.md)
- [AI Search Bar](https://braintrust.dev/docs/cookbook/recipes/AISearch.md)
- [How Zapier uses assertions to evaluate tool usage in chatbots](https://braintrust.dev/docs/cookbook/recipes/Assertions.md)
- [Generating release notes and hill-climbing to improve them](https://braintrust.dev/docs/cookbook/recipes/ReleaseNotes.md)
- [Generating beautiful HTML components](https://braintrust.dev/docs/cookbook/recipes/HTMLGenerator.md)
- [Coda's Help Desk with and without RAG](https://braintrust.dev/docs/cookbook/recipes/CodaHelpDesk.md)
- [Improving Github issue titles using their contents](https://braintrust.dev/docs/cookbook/recipes/Github-Issues.md)
- [Classifying news articles](https://braintrust.dev/docs/cookbook/recipes/ClassifyingNewsArticles.md)
- [Text-to-SQL](https://braintrust.dev/docs/cookbook/recipes/Text2SQL.md)

## All troubleshooting guides

- [Knowledge Base](https://braintrust.dev/docs/kb/index.md): Troubleshooting guides, configuration instructions, and system documentation
- [Troubleshooting guides](https://braintrust.dev/docs/kb/troubleshooting-guides.md): Resolve errors, debug issues, and fix common problems
- [Config guides](https://braintrust.dev/docs/kb/config-guides.md): Set up features, configure integrations, and enable capabilities
- [Process & systems](https://braintrust.dev/docs/kb/process-systems.md): Understand system behavior and technical architecture
- [Azure AD client secret rotation SSO failures](https://braintrust.dev/docs/kb/azure-ad-client-secret-rotation-sso-failures.md)
- [OTLP traces endpoint payload limits](https://braintrust.dev/docs/kb/otlp-traces-endpoint-payload-limits.md)
- [Include runtime LLM output in dataset exports](https://braintrust.dev/docs/kb/include-runtime-llm-output-in-dataset-exports.md)
- [Routing Python SDK logs to multiple Braintrust backends](https://braintrust.dev/docs/kb/python-sdk-byoc-routing-and-state-setup.md)

## Changelog

- [Product updates](https://braintrust.dev/docs/changelog.md): New updates and product improvements
- [Self-hosting releases](https://braintrust.dev/docs/data-plane-changelog.md): Release notes and infrastructure requirements for self-hosted Braintrust deployments

## SDK releases

- [TypeScript SDK changelog](https://braintrust.dev/docs/changelog/typescript-sdk.md)
- [Python SDK changelog](https://braintrust.dev/docs/changelog/python-sdk.md)
- [Java SDK changelog](https://braintrust.dev/docs/changelog/java-sdk.md)
- [Go SDK changelog](https://braintrust.dev/docs/changelog/go-sdk.md)
- [Ruby SDK changelog](https://braintrust.dev/docs/changelog/ruby-sdk.md)
- [C# SDK changelog](https://braintrust.dev/docs/changelog/csharp-sdk.md)

## Other

- [CLAUDE](https://braintrust.dev/docs/CLAUDE.md)
- [500 error when iterating over a dataset](https://braintrust.dev/docs/kb/500-error-when-iterating-over-a-dataset.md)
- [Access UI parameters in remote eval tasks](https://braintrust.dev/docs/kb/access-ui-parameters-in-remote-eval-tasks.md)
- [Accessing log attachments in Python scorers](https://braintrust.dev/docs/kb/accessing-log-attachments-in-python-scorers.md)
- [Add custom measures to monitoring charts](https://braintrust.dev/docs/kb/add-custom-measures-to-monitoring-charts.md)
- [Add tags to scorers using API instead of SDK](https://braintrust.dev/docs/kb/add-tags-to-scorers-using-api-instead-of-sdk.md)
- [Add to dataset captures root span only](https://braintrust.dev/docs/kb/add-to-dataset-captures-root-span-only.md)
- [Aggregate scores across multiple experiments with BTQL](https://braintrust.dev/docs/kb/aggregate-scores-across-multiple-experiments-with-btql.md)
- [AI Proxy audio compression with x-bt-compress-audio](https://braintrust.dev/docs/kb/ai-proxy-audio-compression-with-x.md)
- [Alert aggregation not supported for error rates](https://braintrust.dev/docs/kb/alert-aggregation-not-supported-for-error-rates.md)
- [Alerting on scorer errors and aggregated scores](https://braintrust.dev/docs/kb/alerting-on-scorer-errors-and-aggregated-scores.md)
- [API keys cannot create API keys via impersonation](https://braintrust.dev/docs/kb/api-keys-cannot-create-api-keys-via-impersonation.md)
- [API rate limits during backfill operations](https://braintrust.dev/docs/kb/api-rate-limits-during-backfill-operations.md)
- [API URL configuration conflicts in self-hosted deployments](https://braintrust.dev/docs/kb/api-url-configuration-conflicts-in-self.md)
- [Apply log retention policies across all projects via API](https://braintrust.dev/docs/kb/apply-log-retention-policies-across-all-projects-via-api.md)
- [Assigning log rows to users for review using the API](https://braintrust.dev/docs/kb/assigning-log-rows-to-users-for-review-using-the-api-1780952401983.md)
- [Attaching debuggers to Braintrust eval CLI](https://braintrust.dev/docs/kb/attaching-debuggers-to-braintrust-eval-cli.md)
- [Autoevals Scorer 401 Braintrust gateway error](https://braintrust.dev/docs/kb/autoevals-scorer-401-braintrust-gateway-error.md)
- [Automation scorer failures due to Gemini gateway routing errors](https://braintrust.dev/docs/kb/automation-scorer-failures-due-to-gemini-gateway-routing-errors.md)
- [AWS Bedrock IAM permissions for Braintrust integration](https://braintrust.dev/docs/kb/aws-bedrock-iam-permissions-for-braintrust-integration.md)
- [Bedrock authorization errors with regional access](https://braintrust.dev/docs/kb/bedrock-authorization-errors-with-regional-access.md)
- [Bedrock model variations for on-demand throughput](https://braintrust.dev/docs/kb/bedrock-model-variations-for-on-demand-throughput.md)
- [Bedrock Nova 2 inference-profile IAM and custom provider setup](https://braintrust.dev/docs/kb/bedrock-nova-2-inference-profile-iam-and-custom-provider-setup.md)
- [Blank comparison rows due to mismatched inputs](https://braintrust.dev/docs/kb/blank-comparison-rows-due-to-mismatched-inputs.md)
- [Blank iframe requires console log inspection](https://braintrust.dev/docs/kb/blank-iframe-requires-console-log-inspection.md)
- [Brainstore OOM errors and retry-induced rate limits](https://braintrust.dev/docs/kb/brainstore-oom-errors-and-retry-induced-rate-limits.md)
- [BTQL Filter Not Applied to All Results: Understanding traces vs spans in project_logs Queries](https://braintrust.dev/docs/kb/btql-filter-not-applied-to-all-results-understanding-traces-vs-spans-in-project-logs-queries.md)
- [BTQL multiline query formatting error](https://braintrust.dev/docs/kb/btql-multiline-query-formatting-error.md)
- [BTQL POST endpoint payload and response schema](https://braintrust.dev/docs/kb/btql-post-endpoint-payload-and-response-schema.md)
- [BTQL rate limits on Starter and Pro plans](https://braintrust.dev/docs/kb/btql-rate-limits-on-free-and-pro-plans.md)
- [BTQL summary queries fail with signBlob permission error](https://braintrust.dev/docs/kb/btql-summary-queries-fail-with-signblob-permission-error.md)
- [Chart not visible after creation on monitoring page](https://braintrust.dev/docs/kb/chart-not-visible-after-creation-on-monitoring-page.md)
- [Check data plane version in self-hosted deployments](https://braintrust.dev/docs/kb/check-data-plane-version-in-self.md)
- [Claude Agent SDK API keys recorded in trace metadata](https://braintrust.dev/docs/kb/claude-agent-sdk-api-keys-recorded-in-trace-metadata.md): Understand whether Claude Agent SDK traces contain API keys, how to detect affected spans safely, and how to upgrade and rotate credentials.
- [Claude Code tracing not working after setup](https://braintrust.dev/docs/kb/claude-code-tracing-not-working-after-setup.md)
- [Claude Sonnet 4 missing from playground model picker](https://braintrust.dev/docs/kb/claude-sonnet-4-missing-from-playground-model-picker.md)
- [CLI auth fails for self-hosted deployments with TLS errors](https://braintrust.dev/docs/kb/cli-auth-fails-for-self-hosted-deployments-with-tls-errors.md)
- [CloudFront 403 errors during evaluator invocation](https://braintrust.dev/docs/kb/cloudfront-403-errors-during-evaluator-invocation.md)
- [Code scorer fails with pagination key error](https://braintrust.dev/docs/kb/code-scorer-fails-with-pagination-key-error.md)
- [Code scorer function must be named handler](https://braintrust.dev/docs/kb/code-scorer-function-must-be-named-handler.md)
- [Calculate per-trace averages in Monitor with custom measures](https://braintrust.dev/docs/kb/compute-average-cost-and-llm-calls-per-trace.md)
- [Configure Bedrock with IAM role assumption in Braintrust](https://braintrust.dev/docs/kb/configure-bedrock-with-iam-role-assumption-in-braintrust.md)
- [Configure custom model costs for estimation](https://braintrust.dev/docs/kb/configure-custom-model-costs-for-estimation.md)
- [Configure dataset schemas via API](https://braintrust.dev/docs/kb/configure-dataset-schemas-via-api.md)
- [Configure default groups for Google SSO users](https://braintrust.dev/docs/kb/configure-default-groups-for-google-sso-users.md)
- [Configure human review visibility in dataset views](https://braintrust.dev/docs/kb/configure-human-review-visibility-in-dataset-views.md)
- [Allowlist Braintrust egress IPs for private provider endpoints](https://braintrust.dev/docs/kb/configure-litellm-as-a-custom-provider.md)
- [Configure MCP server with self-hosted data plane](https://braintrust.dev/docs/kb/configure-mcp-server-with-self-hosted-data-plane.md)
- [Configure root span filter for online scoring](https://braintrust.dev/docs/kb/configure-root-span-filter-for-online-scoring.md)
- [Configure scorer pass threshold via API](https://braintrust.dev/docs/kb/configure-scorer-pass-threshold-via-api.md)
- [Context length error caused by wrong AI provider routing](https://braintrust.dev/docs/kb/context-length-error-caused-by-wrong-ai-provider-routing.md)
- [Controlling Concurrency to Prevent Resource Exhaustion](https://braintrust.dev/docs/kb/controlling-concurrency-to-prevent-resource-exhaustion.md)
- [Copy prompts, scorers, and datasets across projects](https://braintrust.dev/docs/kb/copy-prompts-scorers-and-datasets-across-projects.md)
- [COUNT(DISTINCT) in GROUP BY returns row count](https://braintrust.dev/docs/kb/count-distinct-in-group-by-returns-row-count.md)
- [Create a blank monitoring view via API](https://braintrust.dev/docs/kb/create-a-blank-monitoring-view-via-api.md)
- [Create a root span with client-generated conversation IDs](https://braintrust.dev/docs/kb/create-a-root-span-with-client-generated-conversation-ids.md)
- [Create charts with Loop](https://braintrust.dev/docs/kb/create-charts-with-loop.md)
- [Create Slack alerts via the project_automation API](https://braintrust.dev/docs/kb/create-slack-alerts-via-the-project-automation-api.md): Programmatically create Slack alerts for a project using the project_automation API with a POST request and action type "slack".
- [Creating permalinks to log events via API](https://braintrust.dev/docs/kb/creating-permalinks-to-log-events-via-api.md)
- [Creating prompts with encoded custom provider model names](https://braintrust.dev/docs/kb/creating-prompts-with-custom-provider-model-names.md)
- [Creating trace views requires create prompt permission](https://braintrust.dev/docs/kb/creating-trace-views-requires-create-prompt-permission.md)
- [Cross-project scorer automation missing read access](https://braintrust.dev/docs/kb/cross-project-scorer-automation-missing-read-access.md)
- [Cursor Cloud MCP JWT validation error fix](https://braintrust.dev/docs/kb/cursor-cloud-mcp-jwt-validation-error-fix.md)
- [Custom provider base URL causes chat prompt errors](https://braintrust.dev/docs/kb/custom-provider-base-url-causes-chat-prompt-errors.md)
- [Custom scorer auth fails after data plane v2 migration](https://braintrust.dev/docs/kb/custom-scorer-auth-fails-after-data-plane-v2-migration.md)
- [Custom view code cannot use package imports](https://braintrust.dev/docs/kb/custom-view-code-cannot-use-package-imports.md)
- [Customizing experiment names in the SDK](https://braintrust.dev/docs/kb/customizing-experiment-names-in-the-sdk.md)
- [Customizing Monitoring Views: Removing Charts and Managing View Layouts](https://braintrust.dev/docs/kb/customizing-monitoring-views-removing-charts-and-managing-view-layouts.md)
- [Dashboard logs missing due to span name schema error](https://braintrust.dev/docs/kb/dashboard-logs-missing-due-to-span.md)
- [Dataset format for system and user prompts](https://braintrust.dev/docs/kb/dataset-format-for-system-and-user-prompts.md)
- [Rows not attached to a dataset or experiment](https://braintrust.dev/docs/kb/dataset-not-linked-to-experiment-when-using-evalasync-with-parent.md)
- [Dataset search limited to 1000 records](https://braintrust.dev/docs/kb/dataset-search-limited-to-1000-records.md)
- [Debug hanging eval processes with external dependencies](https://braintrust.dev/docs/kb/debug-hanging-eval-processes-with-external.md)
- [Defining response schema in prompts.create()](https://braintrust.dev/docs/kb/defining-response-schema-in-promptscreate.md)
- [Deleting logs via API using correct row IDs](https://braintrust.dev/docs/kb/deleting-logs-via-api-using-correct-row-ids.md)
- [Detecting Remote Eval and CLI Execution Context: SDK Hooks and Environment Variables](https://braintrust.dev/docs/kb/detecting-remote-eval-and-cli-execution-context-sdk-hooks-and-environment-variables.md)
- [Duplicate a prompt in Braintrust](https://braintrust.dev/docs/kb/duplicate-a-prompt-in-braintrust.md)
- [Edit Topics automation sampling rate after setup](https://braintrust.dev/docs/kb/edit-topics-automation-sampling-rate-after-setup.md)
- [EMFILE Error (Too Many Open Files) During Evaluations: Controlling Concurrency to Prevent Resource Exhaustion](https://braintrust.dev/docs/kb/emfile-error-too-many-open-files-during-evaluations-controlling-concurrency-to-prevent-resource-exhaustion.md)
- [Emit traces to multiple projects simultaneously](https://braintrust.dev/docs/kb/emit-traces-to-multiple-projects-simultaneously.md)
- [Unexpected score columns in Logs](https://braintrust.dev/docs/kb/empty-automation-column-persists-after-scorer-deletion.md)
- [Empty traces with flush race conditions](https://braintrust.dev/docs/kb/empty-traces-with-flush-race-conditions.md)
- [Estimated cost missing in logs table with project-level providers](https://braintrust.dev/docs/kb/estimated-cost-missing-in-logs-table-with-project-level-providers.md)
- [Eval timeout errors for initial scorers](https://braintrust.dev/docs/kb/eval-timeout-errors-for-initial-scorers.md)
- [Experiment Fetch Limited to 1000 Events - Export All](https://braintrust.dev/docs/kb/experiment-fetch-limited-to-1000-events.md)
- [Export all experiment events using BTQL pagination](https://braintrust.dev/docs/kb/export-all-experiment-events-using-btql-pagination.md)
- [Export missing child spans](https://braintrust.dev/docs/kb/export-missing-child-spans.md)
- [Extend retention for specific traces using separate projects](https://braintrust.dev/docs/kb/extend-retention-for-specific-traces-using-separate-projects.md)
- [Score logs by Topics facet values with batch scoring](https://braintrust.dev/docs/kb/facets-field-unavailable-in-automation-sql-filters.md)
- [Failed to fetch full trace in data plane](https://braintrust.dev/docs/kb/failed-to-fetch-full-trace-in-data-plane.md)
- [Fetching human review comments via API and BTQL](https://braintrust.dev/docs/kb/fetching-human-review-comments-via-api-and-btql.md)
- [File attachment support across models](https://braintrust.dev/docs/kb/file-attachment-support-across-models.md)
- [Filter logs for automatically converted file attachments](https://braintrust.dev/docs/kb/filter-logs-by-file-attachments.md): Filter your logs to find traces that contain automatically converted file attachments using SQL span filters.
- [Filter logs by tags and scores with ANY_SPAN](https://braintrust.dev/docs/kb/filter-logs-by-tags-and-scores-with-any_span.md)
- [Filter logs for failed LLM calls](https://braintrust.dev/docs/kb/filter-logs-for-failed-llm-calls.md)
- [Model cost missing when using third-party SDKs](https://braintrust.dev/docs/kb/fireworks-cost-missing-when-using-vercel-ai-sdk.md)
- [Fix AI provider auth format and endpoint path after data plane v2](https://braintrust.dev/docs/kb/fix-ai-provider-auth-format-and-endpoint-path-after-data-plane-v2.md)
- [Fix 'No secrets match the org' error](https://braintrust.dev/docs/kb/fix-no-secrets-match-the-org-error.md)
- [Flagging logs and users for review via API](https://braintrust.dev/docs/kb/flagging-logs-and-users-for-review-via-api.md)
- [Flagging users for review via API](https://braintrust.dev/docs/kb/flagging-users-for-review-via-api.md)
- [Gateway calls fail in bt eval --dev mode](https://braintrust.dev/docs/kb/gateway-calls-fail-in-bt-eval-dev-mode.md)
- [GCP checksum mismatch errors after data plane upgrade to v1.1.32](https://braintrust.dev/docs/kb/gcp-checksum-mismatch-after-data-plane-upgrade.md)
- [GCS Brainstore rate limit errors and retries](https://braintrust.dev/docs/kb/gcs-brainstore-rate-limit-errors-and-retries.md)
- [Grouping logs by field in UI requires SQL sandbox](https://braintrust.dev/docs/kb/grouping-logs-by-field-in-ui-requires-sql-sandbox.md)
- [How to configure OpenAI for EU organizations](https://braintrust.dev/docs/kb/how-to-configure-openai-for-eu-organizations-1776438526526.md)
- [How to find and use self-hosted data plane API endpoints](https://braintrust.dev/docs/kb/how-to-find-and-use-self-hosted-data-plane-api-endpoints.md)
- [How to Save Custom Instructions for Loop Using Project](https://braintrust.dev/docs/kb/how-to-save-custom-instructions-for.md)
- [HTTP 421 errors with self-hosted data planes](https://braintrust.dev/docs/kb/http-421-errors-with-self-hosted-data-planes.md)
- [Human review assignee cannot see assigned experiments](https://braintrust.dev/docs/kb/human-review-assignee-cannot-see-assigned-experiments.md)
- [Human review output format: string vs JSON](https://braintrust.dev/docs/kb/human-review-output-format-string-vs-json.md)
- [Human review visibility in experiment row view](https://braintrust.dev/docs/kb/human-review-visibility-in-experiment-row-view.md)
- [Human review with multiple scores in expected field](https://braintrust.dev/docs/kb/human-review-with-multiple-scores-in.md)
- [Ingest OTLP log records into Braintrust](https://braintrust.dev/docs/kb/ingest-otlp-log-records-into-braintrust.md)
- [Internal Server Error: No Secrets Match ORG_NAME in](https://braintrust.dev/docs/kb/internal-server-error-no-secrets-match.md)
- [Java SDK proxy configuration with authentication](https://braintrust.dev/docs/kb/java-sdk-proxy-configuration-with-authentication.md)
- [Keyboard shortcuts for trace and span navigation](https://braintrust.dev/docs/kb/keyboard-shortcuts-for-trace-and-span-navigation.md): Navigate traces and spans in the Braintrust UI using keyboard shortcuts, including arrow keys for the log list and p/n for span navigation.
- [Large experiment downloads timeout in UI](https://braintrust.dev/docs/kb/large-experiment-downloads-timeout-in-ui.md)
- [Large trace download fails with 'Invalid string length'](https://braintrust.dev/docs/kb/large-trace-download-fails-with-invalid-string-length.md)
- [Link prompt versions to traces from code](https://braintrust.dev/docs/kb/link-prompt-versions-to-traces-from-code.md)
- [LiteLLM instrumentation may record provider credentials](https://braintrust.dev/docs/kb/litellm-instrumentation-may-record-provider-credentials.md): Review common questions about the LiteLLM instrumentation security advisory, affected versions, credential exposure, and remediation.
- [loadPrompt caching behavior in TypeScript SDK](https://braintrust.dev/docs/kb/loadprompt-caching-behavior-and-version-fallback.md)
- [Lock down a single project with org groups](https://braintrust.dev/docs/kb/lock-down-a-single-project-with-org-groups.md): Lock down a project by replacing broad organization-wide access with project-specific permission groups.
- [Log filtering with large user ID sets](https://braintrust.dev/docs/kb/log-filtering-with-large-user-id-sets.md)
- [Log filtering with tags stalls after 50 results](https://braintrust.dev/docs/kb/log-filtering-with-tags-stalls-after-50-results.md)
- [Log token metrics to @traced spans manually](https://braintrust.dev/docs/kb/log-token-metrics-to-traced-spans-manually.md)
- [Log user feedback selectively with init_logger](https://braintrust.dev/docs/kb/log-user-feedback-selectively-with-init-logger.md)
- [Logger and experiment span parent type mismatch error](https://braintrust.dev/docs/kb/logger-and-experiment-span-parent-type-mismatch-error.md)
- [Logs not opening due to checksum instability](https://braintrust.dev/docs/kb/logs-not-opening-due-to-checksum-instability.md)
- [Logs page broken by custom column expression](https://braintrust.dev/docs/kb/logs-page-broken-by-custom-column-expression.md): Resolve "Field 'spans' not found in schema" errors caused by a saved custom column expression in the Logs table.
- [Logs stopped after Brainstore platform update](https://braintrust.dev/docs/kb/logs-stopped-after-brainstore-platform-update.md)
- [Loop models require Anthropic API key, not just Bedrock](https://braintrust.dev/docs/kb/loop-models-require-anthropic-api-key-not-just-bedrock.md)
- [Loop security model and user permissions](https://braintrust.dev/docs/kb/loop-security-model-and-user-permissions.md)
- [Enable provider models in Loop](https://braintrust.dev/docs/kb/manage-loop-model-defaults-and-cost.md): Configure the Loop model allowlist so models from your organization's configured AI providers appear in the model picker for Loop sessions.
- [Managing Prompt Updates in Production: Environment Tags](https://braintrust.dev/docs/kb/managing-prompt-updates-in-production-environment.md)
- [Managing your organization's support requests](https://braintrust.dev/docs/kb/managing-your-organizations-support-requests.md)
- [MCP returns empty logs for SaaS org from self-hosted endpoint](https://braintrust.dev/docs/kb/mcp-returns-empty-logs-for-saas-org-from-self-hosted-endpoint.md)
- [Metadata filtering in custom views shows partial data](https://braintrust.dev/docs/kb/metadata-filtering-in-custom-views-shows.md)
- [Minimum OpenAI API key permissions for Braintrust](https://braintrust.dev/docs/kb/minimum-openai-api-key-permissions-for-braintrust.md)
- [Missing classification values from automated scorer](https://braintrust.dev/docs/kb/missing-classification-values-from-automated-scorer.md)
- [Missing scorers in automation rules editor](https://braintrust.dev/docs/kb/missing-scorers-in-automation-rules-editor.md)
- [Moderation AutoEval missing from automation rules scorer picker](https://braintrust.dev/docs/kb/moderation-autoeval-missing-from-automation-rules-scorer-picker.md)
- [Navigate and filter metadata arrays by index](https://braintrust.dev/docs/kb/navigate-and-filter-metadata-arrays-by-index.md)
- [Nested LLM call counting discrepancy](https://braintrust.dev/docs/kb/nested-llm-call-counting-discrepancy.md)
- [No export-only permission in Braintrust ACL model](https://braintrust.dev/docs/kb/no-export-only-permission-in-braintrust-acl-model.md)
- [Node version causing long invoke timeouts](https://braintrust.dev/docs/kb/node-version-causing-long-invoke-timeouts.md)
- [Null input on spans when using OTEL attributes](https://braintrust.dev/docs/kb/null-input-on-spans-when-using-otel-attributes.md)
- [Okta SAML tile login fails with missing RelayState](https://braintrust.dev/docs/kb/okta-saml-tile-login-fails-with-missing-relaystate.md): Troubleshoot the missing RelayState error when signing in to Braintrust from an Okta tile with SAML SSO.
- [Online scorer deduplication with concurrent feedback](https://braintrust.dev/docs/kb/online-scorer-deduplication-with-concurrent-feedback.md)
- [Online scorers not finding production traces](https://braintrust.dev/docs/kb/online-scorers-not-finding-production-traces-time-based-filter-issues-and-query-best-practices.md)
- [Online scoring results not visible in UI](https://braintrust.dev/docs/kb/online-scoring-results-not-visible-in-ui-understanding-async-scoring-and-alert-timing.md)
- [Online scoring rule failing after SDK update](https://braintrust.dev/docs/kb/online-scoring-rule-failing-after-sdk-update.md)
- [Online scoring service accounts in project ACLs](https://braintrust.dev/docs/kb/online-scoring-service-accounts-in-project-acls.md)
- [Open trace URLs in a specific view mode](https://braintrust.dev/docs/kb/open-trace-urls-in-a-specific-view-mode.md)
- [Optimize large trace logging with attachments](https://braintrust.dev/docs/kb/optimize-large-trace-logging-with-attachments.md)
- [Org-wide API key auditing is UI-only](https://braintrust.dev/docs/kb/org-wide-api-key-auditing-is-ui-only.md)
- [OTEL metadata mapping and attachment rendering with PydanticAI](https://braintrust.dev/docs/kb/otel-metadata-mapping-and-attachment-rendering-with-pydanticai.md)
- [Permissions required to view custom trace views](https://braintrust.dev/docs/kb/permissions-required-to-view-custom-trace-views.md)
- [Playground gateway error despite valid API keys](https://braintrust.dev/docs/kb/playground-gateway-error-despite-valid-api-keys.md)
- [Playground timeout with long-running model requests](https://braintrust.dev/docs/kb/playground-timeout-with-long-running-model.md)
- [Populate project names in list prompts API](https://braintrust.dev/docs/kb/populate-project-names-in-list-prompts-api.md)
- [Recover a deleted or inaccessible project](https://braintrust.dev/docs/kb/project-disappeared-or-inaccessible-data-recovery.md): Diagnose why a project disappeared, check whether it was deleted, and request recovery from Braintrust support.
- [Promoting prompts across environments via API](https://braintrust.dev/docs/kb/promoting-prompts-across-environments-via-api-complete-workflow-for-dev-staging-and-production.md)
- [Prompt shows None in experiment view](https://braintrust.dev/docs/kb/prompt-shows-none-in-experiment-view.md)
- [How prompt version IDs work in Braintrust SDKs](https://braintrust.dev/docs/kb/prompt-version-returns-decimal-id-not-ui-hex-string.md)
- [Provision service account access with permission groups](https://braintrust.dev/docs/kb/provision-service-account-access-with-permission-groups.md)
- [Project-level AI provider ignored by legacy proxy](https://braintrust.dev/docs/kb/proxy-provider-precedence-and-routing-guide.md)
- [PydanticAI integration counts wrapper spans as LLM calls](https://braintrust.dev/docs/kb/pydanticai-integration-counts-wrapper-spans-as.md)
- [Configure a custom CA for Python scorers on self-hosted AWS](https://braintrust.dev/docs/kb/python-scorer-custom-ca-bundle-aws-kb.md): Package a custom CA bundle so Python scorers in AWS Lambda quarantine can complete Braintrust SDK login through TLS-intercepting proxies before scorer code runs.
- [Python SDK Docker containerization setup](https://braintrust.dev/docs/kb/python-sdk-docker-containerization-setup.md)
- [Python sdk s3 presigned response handling](https://braintrust.dev/docs/kb/python-sdk-s3-presigned-response-handling.md)
- [Query multiple projects for combined metrics](https://braintrust.dev/docs/kb/query-multiple-projects-for-combined-metrics.md)
- [Query scorer span token costs with SQL](https://braintrust.dev/docs/kb/query-scorer-span-token-costs-with-sql.md)
- [Query timeout requires timestamp filter](https://braintrust.dev/docs/kb/query-timeout-requires-timestamp-filter.md)
- [Query token distribution by user or prompt with SQL](https://braintrust.dev/docs/kb/query-token-distribution-by-user-or-prompt-with-sql.md)
- [Rate limit error during dataset iteration](https://braintrust.dev/docs/kb/rate-limit-error-during-dataset-iteration.md)
- [Re-running scorers without re-running tasks](https://braintrust.dev/docs/kb/re-running-scorers-without-re-running-tasks.md)
- [Recover or retrieve a deleted prompt](https://braintrust.dev/docs/kb/recover-or-retrieve-a-deleted-prompt.md)
- [Recovering deleted experiment rows with versioned BTQL queries](https://braintrust.dev/docs/kb/recovering-deleted-experiment-rows.md)
- [Remote eval dev mode fails with Pydantic parameters](https://braintrust.dev/docs/kb/remote-eval-dev-mode-fails-with-pydantic-parameters.md)
- [Remote eval OpenTelemetry header propagation](https://braintrust.dev/docs/kb/remote-eval-opentelemetry-header-propagation.md): Propagate OpenTelemetry trace context from a remote eval handler by creating spans and attaching Braintrust request context as span attributes.
- [Removing charts from the default "All data view"](https://braintrust.dev/docs/kb/removing-charts-from-the-default-all-data-view.md)
- [Repair unhealthy Topics automation references](https://braintrust.dev/docs/kb/repair-unhealthy-topics-automation-references.md): Identify missing facet or topic map references, repair unhealthy Topics automation rows, and collect diagnostics for support.
- [Reporting a critical outage](https://braintrust.dev/docs/kb/reporting-a-critical-outage-1778110794263.md)
- [Fix “slug already exists” errors for prompts that are not visible](https://braintrust.dev/docs/kb/resolve-invisible-prompt-slug-conflict-with-put.md)
- [Restrict user access to specific projects only](https://braintrust.dev/docs/kb/restrict-user-access-to-specific-projects-only.md)
- [Retry-safe eval workers with stable row IDs](https://braintrust.dev/docs/kb/retry-safe-eval-workers-with-stable-row-ids.md)
- [Review specific rows in Logs and Experiments](https://braintrust.dev/docs/kb/review-button-ignores-selected-rows-in-logs.md)
- [Root span id encoding for trace grouping](https://braintrust.dev/docs/kb/root-span-id-encoding-for-trace-grouping.md)
- [Route requests to specific AI providers](https://braintrust.dev/docs/kb/route-requests-to-specific-ai-providers.md)
- [Running Evaluations Per Git Commit: SHA-Based Experiment](https://braintrust.dev/docs/kb/running-evaluations-per-git-commit-sha.md)
- [SAML login fails with email domain mismatch](https://braintrust.dev/docs/kb/saml-login-fails-with-email-domain-mismatch.md)
- [Scale S3 export automation across multiple projects](https://braintrust.dev/docs/kb/scale-s3-export-automation-across-multiple-projects.md)
- [Score multi-turn agent conversations across traces](https://braintrust.dev/docs/kb/score-multi-turn-agent-conversations-across-traces.md)
- [Scorer overwriting human review data with identical names](https://braintrust.dev/docs/kb/scorer-overwriting-human-review-data-with.md)
- [Scorer pane columns collapse on open](https://braintrust.dev/docs/kb/scorer-pane-columns-collapse-on-open.md)
- [Scorer template variable reference](https://braintrust.dev/docs/kb/scorer-template-variable-reference-1777485051181.md)
- [Scores added via API not displaying in UI](https://braintrust.dev/docs/kb/scores-added-via-api-not-displaying-in-ui.md)
- [SDK rate limits during dataset iteration](https://braintrust.dev/docs/kb/sdk-rate-limits-during-dataset-iteration.md)
- [Search Timeout on Log Queries: CAST Operations in Custom](https://braintrust.dev/docs/kb/search-timeout-on-log-queries-cast.md)
- [Securing remote eval endpoints: authentication and](https://braintrust.dev/docs/kb/securing-remote-eval-endpoints-authentication-and.md)
- [Service token permissions for log ingestion](https://braintrust.dev/docs/kb/service-token-permissions-for-log-ingestion.md)
- [Set sub-5-minute export automation intervals via API](https://braintrust.dev/docs/kb/set-sub-5-minute-export-automation-intervals-via-api.md)
- [Setting up SSO integration in Braintrust](https://braintrust.dev/docs/kb/setting-up-sso-integration-in-braintrust-1777472841602.md)
- [Setup claude code with OTel tracing](https://braintrust.dev/docs/kb/setup-claude-code-with-otel-tracing-1781641599984.md)
- [Slack alert permissions in project settings](https://braintrust.dev/docs/kb/slack-alert-permissions-in-project-settings.md)
- [Span components v4 encoding error on self-hosted after upgrading the SDK](https://braintrust.dev/docs/kb/span-components-v4-encoding-error-on-self-hosted-after-sdk-upgrade.md): Resolve "SpanComponents string is not properly encoded" errors on self-hosted deployments caused by upgrading the SDK before the Braintrust data plane.
- [Span Name Filtering Not Working on Monitoring Page:](https://braintrust.dev/docs/kb/span-name-filtering-not-working-on.md)
- [SpanComponents encoding error when attaching feedback spans](https://braintrust.dev/docs/kb/spancomponents-encoding-error-when-attaching-feedback-spans.md)
- [SQL queries timeout when missing timestamp filter](https://braintrust.dev/docs/kb/sql-queries-timeout-when-missing-timestamp-filter.md)
- [SSL certificate verification failure behind corporate proxy](https://braintrust.dev/docs/kb/ssl-certificate-verification-failure-behind-corporate-proxy.md)
- [SSO login fails with 'did not grant access' error](https://braintrust.dev/docs/kb/sso-login-fails-with-did-not-grant-access-error.md)
- [Stitch multi-turn chat into one trace with wrapAISDK](https://braintrust.dev/docs/kb/stitch-multi-turn-chat-into-one-trace-with-wrapaisdk.md)
- [Strict mode omits required field in structured output](https://braintrust.dev/docs/kb/strict-mode-omits-required-field-in-structured-output.md)
- [Strip OTel attributes from metadata](https://braintrust.dev/docs/kb/strip-otel-attributes-from-metadata.md)
- [Structured outputs unsupported for Azure providers](https://braintrust.dev/docs/kb/structured-outputs-unsupported-for-azure-providers.md)
- [Tag filter dropdown missing tags from logged data](https://braintrust.dev/docs/kb/tag-filter-dropdown-missing-tags-from-logged-data.md)
- [Temperature parameter included in GPT-5 prompts from SDK](https://braintrust.dev/docs/kb/temperature-parameter-included-in-gpt-5.md)
- [Temperature parameter not supported in GPT-5 models](https://braintrust.dev/docs/kb/temperature-parameter-not-supported-in-gpt.md)
- [Thread view span extraction and filtering](https://braintrust.dev/docs/kb/thread-view-span-extraction-and-filtering.md)
- [Timestamps converted to UTC in traces](https://braintrust.dev/docs/kb/timestamps-converted-to-utc-in-traces.md)
- [Topics automation can exhaust your monthly Topics credit quickly](https://braintrust.dev/docs/kb/topics-automation-exhausts-log-analysis-limits-quickly.md)
- [Topics automation idle timeout and trace update behavior](https://braintrust.dev/docs/kb/topics-automation-idle-timeout-and-trace-update-behavior.md)
- [Topics facets empty due to Skipped status](https://braintrust.dev/docs/kb/topics-facets-empty-due-to-skipped-status.md)
- [Total LLM cost preset shows no data](https://braintrust.dev/docs/kb/total-llm-cost-preset-requirements.md)
- [Trace gaps when exported span context is not persisted across restarts](https://braintrust.dev/docs/kb/trace-gaps-when-parent-span-export-is-not-persisted-across-restarts.md)
- [Tracing Anthropic models via LiteLLM with Braintrust](https://braintrust.dev/docs/kb/tracing-anthropic-models-via-litellm-with-braintrust.md)
- [Tracing OpenAI Realtime API with WebSockets](https://braintrust.dev/docs/kb/tracing-openai-realtime-api-with-websockets.md)
- [Transient JSONL.GZ files in S3 export bucket](https://braintrust.dev/docs/kb/transient-jsonl-gz-files-in-s3-export-bucket.md)
- [Trial quota limits and monthly reset cycle](https://braintrust.dev/docs/kb/trial-quota-limits-and-monthly-reset-cycle.md)
- [Troubleshooting 500 errors in self-hosted data planes](https://braintrust.dev/docs/kb/troubleshooting-500-errors-in-self-hosted.md)
- [Troubleshooting Organization Changes in Self-Hosted Data](https://braintrust.dev/docs/kb/troubleshooting-organization-changes-in-self-hosted.md)
- [Troubleshooting Traces Stuck in Progress](https://braintrust.dev/docs/kb/troubleshooting-traces-stuck-in-progress.md)
- [TypeScript SDK eval process hangs after completion](https://braintrust.dev/docs/kb/typescript-sdk-eval-process-hangs-after-completion-debugging-open-handles-and-resource-leaks.md)
- [UI cache hit rate shows zero with raw metrics](https://braintrust.dev/docs/kb/ui-cache-hit-rate-shows-zero-with-raw-metrics.md): Fix a 0% cache hit rate in the UI by mapping provider cache token fields to the normalized metrics Braintrust expects from custom logging.
- [UI experiment timeout configuration for self-hosted k8s](https://braintrust.dev/docs/kb/ui-experiment-timeout-configuration-for-self-hosted-k8s.md)
- [Understanding Brainstore query optimization warning logs](https://braintrust.dev/docs/kb/understanding-brainstore-query-optimization-warning-logs.md)
- [Understanding Experiment Score Aggregation: Simple vs](https://braintrust.dev/docs/kb/understanding-experiment-score-aggregation-simple-vs.md)
- [Understanding traces vs spans in SQL & BTQL queries](https://braintrust.dev/docs/kb/understanding-traces-vs-spans-in-sql-btql-queries.md)
- [Subquery alias conflicts with column names](https://braintrust.dev/docs/kb/unpivot-subquery-alias-conflicts-with-column-names.md)
- [Update AWS Bedrock credentials using API](https://braintrust.dev/docs/kb/update-aws-bedrock-credentials-using-api.md)
- [Use project IDs to prevent duplicate projects after rename](https://braintrust.dev/docs/kb/use-project-ids-to-prevent-duplicate-projects-after-rename.md)
- [Use S3 export instead of polling SQL for pipelines](https://braintrust.dev/docs/kb/use-s3-export-instead-of-polling-btql-for-pipelines.md)
- [Use `_xact_id` to dedupe exports and determine update time](https://braintrust.dev/docs/kb/use-xact-id-to-dedupe-exports-and-determine-update-time.md): Deduplicate cloud export rows and track updated records using stable IDs and _xact_id.
- [User Engagement Tracking](https://braintrust.dev/docs/kb/user-engagement-tracking.md)
- [Using Google OAuth for Google Workspace domains](https://braintrust.dev/docs/kb/using-google-oauth-for-google-workspace-domains.md): Enable Google Workspace users to sign in with Google OAuth immediately, then migrate to custom SAML SSO when a super-admin is available.
- [Using OpenAI Responses API with Braintrust SDKs](https://braintrust.dev/docs/kb/using-openai-responses-api-with-braintrust-sdks.md)
- [Vertex AI gemini model region and registry issues](https://braintrust.dev/docs/kb/vertex-ai-gemini-model-region-and-registry-issues.md)
- [View logs timeouts from broad IS NOT NULL filters](https://braintrust.dev/docs/kb/view-logs-timeouts-from-broad-is-not-null-filters.md)
- [Configure sustained score alerts with external aggregation](https://braintrust.dev/docs/kb/webhook-based-sustained-alerts-for-llm-scores.md): Alert when LLM score degradations persist over a time window using webhook-based external aggregation or filter approximations.
- [Why mark complete is missing on log span reviews](https://braintrust.dev/docs/kb/why-mark-complete-is-missing-on-log-span-reviews.md)
- [wrapAISDK interleaved traces with concurrent calls](https://braintrust.dev/docs/kb/wrapaisdk-interleaved-traces-with-concurrent-calls.md)
- [Autoevals](https://braintrust.dev/docs/sdks/python/related/autoevals/overview.md): Autoevals is a tool to quickly and easily evaluate AI model outputs
- [Autoevals](https://braintrust.dev/docs/sdks/typescript/related/autoevals/overview.md): Autoevals is a tool to quickly and easily evaluate AI model outputs

## OpenAPI Specs

- [openapi](/docs/openapi.yaml)
- [openapi](/docs/api-reference/openapi.json)
- [openapi](/docs/openapi.json)

> The links below point to documentation indexes. Follow each `/_llms/` index recursively until you reach documentation pages.

## Indexes

- [Reference (254 pages)](https://braintrust.dev/docs/_llms/reference.md): Documentation for Reference.