# Latitude Docs

## Docs

- [Claude Code agent dispatch](https://docs.latitude.so/agent-dispatch/claude-code.md): Trigger Claude Code routines from Latitude signals.
- [Cursor agent dispatch](https://docs.latitude.so/agent-dispatch/cursor.md): Send Latitude signals to Cursor agents so they can investigate issues and open pull requests.
- [Linear agent dispatch](https://docs.latitude.so/agent-dispatch/linear.md): Create Linear issues automatically from Latitude signals.
- [Agent dispatch](https://docs.latitude.so/agent-dispatch/overview.md): Wake a coding agent when a signal needs engineering follow-up, and let it investigate over Latitude's MCP.
- [Agent dispatch webhooks](https://docs.latitude.so/agent-dispatch/webhooks.md): Receive Latitude agent dispatches in your own system.
- [Flaggers](https://docs.latitude.so/annotations/flaggers.md): Automatic annotators that surface common failure categories on every trace
- [Annotate traces effectively](https://docs.latitude.so/annotations/guides/annotate-effectively.md): Practical habits for human annotation that keep signal discovery and evaluation alignment accurate.
- [Inline Annotations](https://docs.latitude.so/annotations/inline-annotations.md): Annotate any trace directly from its detail view
- [Annotations Overview](https://docs.latitude.so/annotations/overview.md): Review your agent's interactions and provide human feedback
- [Add traces to a dataset](https://docs.latitude.so/datasets/add-traces.md): Build a test set from real production traces by adding them to a dataset from the trace list, search results, or a signal.
- [Custom columns](https://docs.latitude.so/datasets/custom-columns.md): Add your own columns to a dataset, rename or reorder them, and remove columns you no longer need without losing data.
- [Edit row contents](https://docs.latitude.so/datasets/edit-rows.md): Update the cells of an existing dataset row — over the API, the SDKs, or an MCP agent — to fill in expected output or correct any field after a row was created.
- [Add expected output](https://docs.latitude.so/datasets/expected-output.md): Record the correct answer for dataset rows so regression tests can check your agent against a known-good result.
- [Datasets](https://docs.latitude.so/datasets/overview.md): Curate collections of inputs, outputs, and expected outputs from real traces to test and improve your agent.
- [Cluster](https://docs.latitude.so/deployment/cluster.md): Deploy a scalable, highly-available Latitude on any Kubernetes cluster with our Helm chart.
- [Configuration](https://docs.latitude.so/deployment/configuration.md): The full Latitude configurable environment-variable reference for self-hosting.
- [Overview](https://docs.latitude.so/deployment/overview.md): Use Latitude as a fully-managed cloud service, or self-host it — from a single machine to a Kubernetes cluster.
- [Single-host](https://docs.latitude.so/deployment/single-host.md): Run a production-grade Latitude instance on a single machine with Docker Compose or Docker Swarm.
- [Contributing](https://docs.latitude.so/development/contributing.md): How to contribute to Latitude — set up locally, propose changes, and follow our pull-request conventions and Code of Conduct.
- [Setup](https://docs.latitude.so/development/setup.md): Run the full Latitude stack on your machine ready for development with hot reload and infrastructure in Docker.
- [Evaluation alignment](https://docs.latitude.so/evaluations/alignment.md): Measure how well your automated evaluations agree with human judgment.
- [Custom scripts](https://docs.latitude.so/evaluations/custom-scripts.md): Write a JavaScript detector that reads a session and decides whether a behavior is present.
- [Detection methods](https://docs.latitude.so/evaluations/detection-methods.md): The three ways an evaluation decides whether a session matches a signal: a set of conditions, an LLM judge, or a custom script.
- [Evaluations overview](https://docs.latitude.so/evaluations/overview.md): Understand how evaluations score traffic to monitor your agent's quality over time.
- [Evaluation triggers](https://docs.latitude.so/evaluations/triggers.md): Configure which sessions an evaluation runs on, and how many.
- [Experiments](https://docs.latitude.so/experiments/overview.md): Compare two or more slices of your data side by side — every metric across sessions, users, tools, signals, behaviours and more.
- [CLI](https://docs.latitude.so/getting-started/cli.md): Install and use the Latitude CLI to manage your Latitude organization right from the terminal
- [Core Concepts](https://docs.latitude.so/getting-started/concepts.md): Latitude's core concepts: projects, spans, traces, sessions, users, tools, search, behaviours, scores, signals, evaluations, and monitors.
- [How to use Latitude](https://docs.latitude.so/getting-started/how-to-use-latitude.md): The recommended workflow for finding, tracking, and fixing production failures in your AI agent.
- [Latitude Overview](https://docs.latitude.so/getting-started/introduction.md): Understand what Latitude is, why teams use its MIT-licensed open-source platform, and how Observe, Understand, and Refine fit together.
- [MCP Server](https://docs.latitude.so/getting-started/mcp.md): Connect AI agents like Claude, Cursor, Codex, and Zed to your Latitude organization through the Latitude MCP server
- [Self-Healing Agents](https://docs.latitude.so/getting-started/self-healing-agents.md): Close the loop from detected production failure to reviewed pull request by dispatching your coding agent with the evidence attached.
- [Skills](https://docs.latitude.so/getting-started/skills.md): Give your agent knowledge on how to set up and manage Latitude autonomously with our skills.
- [Monitors Overview](https://docs.latitude.so/monitors/overview.md): Watch your signals and saved searches, and get alerted when something needs your attention.
- [API Reference](https://docs.latitude.so/more/api-reference.md): Open the Latitude API reference.
- [Overview](https://docs.latitude.so/more/data-destinations/overview.md): Continuously sync your Latitude data into the systems you already use.
- [PostHog](https://docs.latitude.so/more/data-destinations/posthog.md): Send Latitude spans to PostHog LLM Analytics as native AI events.
- [Slack integration](https://docs.latitude.so/more/slack.md): Send Latitude notifications to Slack channels.
- [Duration](https://docs.latitude.so/observability/features/duration.md): Track trace and session duration to investigate latency and long-running conversations.
- [Environments](https://docs.latitude.so/observability/features/environments.md): Separate production, staging, preview, and local traces with tags or metadata.
- [Log levels](https://docs.latitude.so/observability/features/log-levels.md): Track application-defined severity with metadata and combine it with trace status filters.
- [Metadata](https://docs.latitude.so/observability/features/metadata.md): Attach structured key-value context to traces for filtering and investigation.
- [Percentile cohorts](https://docs.latitude.so/observability/features/percentile-cohorts.md): Compare trace duration, TTFT, cost, and token count against similar traces with the same tag set.
- [Releases and versioning](https://docs.latitude.so/observability/features/releases-versioning.md): Track agent versions, deployment ids, and rollout cohorts with tags or metadata.
- [Sampling](https://docs.latitude.so/observability/features/sampling.md): Understand where Latitude does and does not sample trace processing.
- [Sandbox](https://docs.latitude.so/observability/features/sandbox.md): An isolated, free environment to debug your agents with Latitude's traces UI from your dev and staging environments — without polluting production.
- [Tags](https://docs.latitude.so/observability/features/tags.md): Add lightweight labels to traces so you can filter and search cohorts.
- [Token and cost tracking](https://docs.latitude.so/observability/features/token-cost-tracking.md): Track token usage and estimated LLM cost across traces and sessions.
- [Trace IDs](https://docs.latitude.so/observability/features/trace-ids.md): Use OpenTelemetry trace ids to connect Latitude traces with the rest of your observability stack.
- [Trace URLs](https://docs.latitude.so/observability/features/trace-urls.md): Open and share specific traces from the Latitude web app.
- [User tracking](https://docs.latitude.so/observability/features/user-tracking.md): Group and filter traces by the end user that triggered them.
- [Filters](https://docs.latitude.so/observability/filters.md): Use the shared filter system to narrow traces, save searches, and configure evaluation triggers
- [Build a dashboard from Latitude data](https://docs.latitude.so/observability/guides/build-a-dashboard.md): Give an agent arbitrarily sliceable analytics and let it render a self-contained HTML dashboard.
- [Group traces by project](https://docs.latitude.so/observability/guides/group-traces-by-project.md): Keep signal detection scoped to the right agent by routing traces into separate Latitude projects.
- [Memory](https://docs.latitude.so/observability/memory.md): Watch how your agents' persistent memory evolves, with per-record history, diffs, and the trace behind every change.
- [Observability overview](https://docs.latitude.so/observability/overview.md): Inspect agent traces, spans, sessions, and telemetry metadata in Latitude.
- [Sessions](https://docs.latitude.so/observability/sessions.md): Group related traces into multi-turn conversations.
- [Spans](https://docs.latitude.so/observability/spans.md): A span is the smallest unit of captured work inside a trace, such as an LLM call, a tool call, or a retrieval step.
- [Tool calls](https://docs.latitude.so/observability/tool-calls.md): Inspect the tools your agent invoked, including their inputs, outputs, and errors, as spans inside a trace.
- [Tools](https://docs.latitude.so/observability/tools.md): Monitor every tool your agents define and call, with usage, failure, and latency metrics.
- [Traces](https://docs.latitude.so/observability/traces.md): Understand the trace model, lifecycle, and how to work with traces in Latitude
- [Users](https://docs.latitude.so/observability/users.md): Review activity, sessions, and errors by end user.
- [Score Analytics](https://docs.latitude.so/scores/analytics.md): Visualize score trends and quality metrics across your project
- [Scores Overview](https://docs.latitude.so/scores/overview.md): Understand how scores work as the universal measurement unit in Latitude
- [Behaviours](https://docs.latitude.so/search/behaviours.md): Browse the topics Latitude automatically discovers across your sessions, with trends and outcome metrics, to find what matters without writing a search.
- [Search and review effectively](https://docs.latitude.so/search/guides/search-and-review-effectively.md): Practical habits for writing queries, saving useful searches, and keeping them useful over time.
- [Search Overview](https://docs.latitude.so/search/overview.md): Find traces by meaning or exact text, combine search with filters, and save the searches your team uses repeatedly
- [Saved Searches](https://docs.latitude.so/search/saved-searches.md): Bookmark useful search queries and filter sets, then reopen them from the Saved searches dropdown
- [GDPR](https://docs.latitude.so/security/compliance/gdpr.md): Latitude's GDPR readiness, AWS eu-central-1 data boundary, and customer data protection practices.
- [ISO 27001](https://docs.latitude.so/security/compliance/iso-27001.md): Latitude's ISO 27001 audit status and information security management program.
- [SOC II](https://docs.latitude.so/security/compliance/soc2.md): Latitude's SOC II certification and security controls.
- [Data protection](https://docs.latitude.so/security/data-protection.md): How Latitude keeps hosted customer data and Latitude-managed inference within AWS eu-central-1.
- [PII redaction](https://docs.latitude.so/security/pii-redaction.md): How Latitude reduces PII exposure with SDK redaction and internal data handling safeguards.
- [Create a signal](https://docs.latitude.so/signals/create.md): Define a signal yourself when you already know the behavior you want to track, using the New signal builder.
- [Signal management](https://docs.latitude.so/signals/management.md): Triage signals, monitor them with evaluations, resolve what you fix, ignore noise, and keep your signal list focused.
- [Signal discovery](https://docs.latitude.so/signals/overview.md): Understand how Latitude turns annotations, scores, evaluations, and flaggers into trackable production signals.
- [Claude code](https://docs.latitude.so/telemetry/claude-code.md)
- [Cloudflare AI Gateway](https://docs.latitude.so/telemetry/frameworks/cloudflare-ai-gateway.md): Connect Cloudflare AI Gateway to Latitude for observability.
- [Cloudflare Think](https://docs.latitude.so/telemetry/frameworks/cloudflare-think.md): Connect Cloudflare Think agents to Latitude for observability.
- [CrewAI](https://docs.latitude.so/telemetry/frameworks/crewai.md): Connect your CrewAI multi-agent application to Latitude for observability.
- [DSPy](https://docs.latitude.so/telemetry/frameworks/dspy.md): Connect your DSPy program to Latitude for observability.
- [ElevenLabs Agents](https://docs.latitude.so/telemetry/frameworks/elevenlabs.md): Connect your ElevenLabs voice agents to Latitude for LLM observability.
- [Eve](https://docs.latitude.so/telemetry/frameworks/eve.md): Connect your Eve-powered agent to Latitude for observability.
- [Flue](https://docs.latitude.so/telemetry/frameworks/flue.md): Connect your Flue workflows and agents to Latitude for observability.
- [Google ADK](https://docs.latitude.so/telemetry/frameworks/google-adk.md): Connect your Google Agent Development Kit (ADK) application to Latitude for observability.
- [Haystack](https://docs.latitude.so/telemetry/frameworks/haystack.md): Connect your Haystack application to Latitude for observability.
- [LangChain](https://docs.latitude.so/telemetry/frameworks/langchain.md): Connect your LangChain-powered application to Latitude for observability.
- [LiteLLM](https://docs.latitude.so/telemetry/frameworks/litellm.md): Connect your LiteLLM-powered application to Latitude for observability.
- [LiveKit Agents](https://docs.latitude.so/telemetry/frameworks/livekit.md): Connect your LiveKit voice AI agents to Latitude for observability.
- [LlamaIndex](https://docs.latitude.so/telemetry/frameworks/llamaindex.md): Connect your LlamaIndex-powered application to Latitude for observability.
- [Mastra](https://docs.latitude.so/telemetry/frameworks/mastra.md): Connect your Mastra-powered application to Latitude for observability.
- [OpenAI Agents SDK](https://docs.latitude.so/telemetry/frameworks/openai-agents.md): Connect your OpenAI Agents SDK application to Latitude for observability.
- [Pydantic AI](https://docs.latitude.so/telemetry/frameworks/pydantic-ai.md): Connect your Pydantic AI agent to Latitude for observability.
- [Strands Agents](https://docs.latitude.so/telemetry/frameworks/strands.md): Connect your Strands Agents application to Latitude for observability.
- [Vercel AI SDK](https://docs.latitude.so/telemetry/frameworks/vercel-ai-sdk.md): Connect your Vercel AI SDK-powered application to Latitude for observability.
- [Vercel AI SDK v7](https://docs.latitude.so/telemetry/frameworks/vercel-ai-sdk-v7.md): Connect your Vercel AI SDK v7 powered application to Latitude for observability.
- [Hermes](https://docs.latitude.so/telemetry/hermes.md)
- [Memory tracing](https://docs.latitude.so/telemetry/memory.md): Instrument your agent's long-term memory so Latitude can show how it evolves.
- [Openclaw](https://docs.latitude.so/telemetry/openclaw.md)
- [OpenTelemetry Exporter (OTEL)](https://docs.latitude.so/telemetry/otel-exporter.md): Connect any OpenTelemetry-instrumented application to Latitude, regardless of language or framework.
- [Pi coding agent](https://docs.latitude.so/telemetry/pi-coding-agent.md)
- [Prime intellect](https://docs.latitude.so/telemetry/prime-intellect.md)
- [Aleph Alpha](https://docs.latitude.so/telemetry/providers/aleph-alpha.md): Connect your Aleph Alpha-powered application to Latitude for observability.
- [Amazon Bedrock](https://docs.latitude.so/telemetry/providers/amazon-bedrock.md): Connect your Amazon Bedrock-powered application to Latitude for observability.
- [Anthropic](https://docs.latitude.so/telemetry/providers/anthropic.md): Connect your Anthropic-powered application to Latitude for observability.
- [Azure OpenAI](https://docs.latitude.so/telemetry/providers/azure.md): Connect your Azure OpenAI-powered application to Latitude for observability.
- [Cohere](https://docs.latitude.so/telemetry/providers/cohere.md): Connect your Cohere-powered application to Latitude for observability.
- [Google Gemini](https://docs.latitude.so/telemetry/providers/gemini.md): Connect your Google Gemini (Developer API) application to Latitude for observability.
- [Google AI Platform](https://docs.latitude.so/telemetry/providers/google-ai-platform.md): Connect your Google AI Platform-powered application to Latitude for observability.
- [Groq](https://docs.latitude.so/telemetry/providers/groq.md): Connect your Groq-powered application to Latitude for observability.
- [Mistral AI](https://docs.latitude.so/telemetry/providers/mistral.md): Connect your Mistral AI-powered application to Latitude for observability.
- [Ollama](https://docs.latitude.so/telemetry/providers/ollama.md): Connect your Ollama-powered application to Latitude for observability.
- [OpenAI](https://docs.latitude.so/telemetry/providers/openai.md): Connect your OpenAI-powered application to Latitude for observability.
- [Replicate](https://docs.latitude.so/telemetry/providers/replicate.md): Connect your Replicate-powered application to Latitude for observability.
- [Amazon SageMaker](https://docs.latitude.so/telemetry/providers/sagemaker.md): Connect your Amazon SageMaker-powered application to Latitude for observability.
- [Together AI](https://docs.latitude.so/telemetry/providers/together-ai.md): Connect your Together AI-powered application to Latitude for observability.
- [Hugging Face Transformers](https://docs.latitude.so/telemetry/providers/transformers.md): Connect your Hugging Face Transformers application to Latitude for observability.
- [Vertex AI](https://docs.latitude.so/telemetry/providers/vertex-ai.md): Connect your Vertex AI-powered application to Latitude for observability.
- [IBM watsonx.ai](https://docs.latitude.so/telemetry/providers/watsonx.md): Connect your IBM watsonx.ai-powered application to Latitude for observability.
- [Python SDK](https://docs.latitude.so/telemetry/python.md): Instrument Python apps with Latitude Telemetry.
- [Start tracing](https://docs.latitude.so/telemetry/start-tracing.md): Connect your agent to Latitude and send your first traces.
- [TypeScript SDK](https://docs.latitude.so/telemetry/typescript.md): Instrument TypeScript and JavaScript apps with Latitude Telemetry.
- [Regression testing](https://docs.latitude.so/test-and-fix/regression-testing.md): Replay a dataset of real traces against your agent and check the results with the same evaluations that monitor production, so fixed failures cannot quietly return.

## Optional

- [GitHub](https://github.com/latitude-dev/latitude-llm)
- [YouTube](https://www.youtube.com/@trylatitude)
- [Slack Community](https://join.slack.com/t/trylatitude/shared_invite/zt-35wu2h9es-N419qlptPMhyOeIpj3vjzw)