# AgentDbg

## Docs

- [Configure AgentDbg: Env Vars, YAML, and Precedence](https://refinehq.ai/docs/configuration/overview.md): AgentDbg reads settings from env vars and two YAML config files. Env vars win over project YAML, which wins over user YAML, then built-in defaults.
- [Redact Secrets from Agent Traces Automatically](https://refinehq.ai/docs/configuration/redaction.md): AgentDbg scrubs values for sensitive keys before writing any trace data to disk, keeping API keys and tokens out of your local trace files by default.
- [Local Trace Storage, File Layout, and Run Naming](https://refinehq.ai/docs/configuration/storage.md): AgentDbg stores every trace as plain JSON files on your machine. Learn where files land, how to move the data directory, and how to name and delete runs.
- [Debug CrewAI Agents and Flows Locally with AgentDbg](https://refinehq.ai/docs/integrations/crewai.md): Import the AgentDbg CrewAI adapter to automatically record every LLM call and tool call from your crews and flows into a local AgentDbg timeline.
- [Auto-Instrument LangChain Agents with AgentDbg](https://refinehq.ai/docs/integrations/langchain.md): Use AgentDbgLangChainCallbackHandler to automatically capture every LLM and tool call from your LangChain or LangGraph agent into an AgentDbg trace.
- [Trace OpenAI Agents SDK Runs Locally with AgentDbg](https://refinehq.ai/docs/integrations/openai-agents.md): Register the AgentDbg tracing processor to forward OpenAI Agents SDK generation, function, and handoff spans into a local AgentDbg timeline automatically.
- [AgentDbg: The Step-Through Debugger for AI Agents](https://refinehq.ai/docs/introduction.md): Learn what AgentDbg is, what problems it solves, and what gets captured: LLM calls, tool calls, state updates, errors, and loop warnings.
- [Quickstart: Trace Your First Agent with AgentDbg](https://refinehq.ai/docs/quickstart.md): Install AgentDbg, instrument your first agent with @trace and the recording helpers, run it, and open the timeline UI—all in under 5 minutes.
- [AgentDbg CLI Reference: list, view, and export](https://refinehq.ai/docs/reference/cli.md): Complete reference for the agentdbg command-line tool. Covers list, view, and export commands with all flags: --limit, --json, --no-browser, and --port.
- [AgentDbg Trace File Format Reference (spec v0.1)](https://refinehq.ai/docs/reference/trace-format.md): Reference for run.json and events.jsonl files AgentDbg writes to disk. Covers all event types, payload schemas, field types, and the spec_version contract.
- [Using the AgentDbg Timeline Viewer to Inspect Runs](https://refinehq.ai/docs/reference/viewer.md): The AgentDbg timeline viewer shows every run event in your browser. Navigate, filter, rename, and delete runs directly from the local UI—no cloud required.
- [Stop Runaway Agents with AgentDbg Run Guardrails](https://refinehq.ai/docs/sdk/guardrails.md): Configure AgentDbg guardrails to automatically abort agent runs that loop indefinitely, exceed LLM call budgets, or run longer than expected.
- [Record LLM Calls, Tool Calls, and State in AgentDbg](https://refinehq.ai/docs/sdk/recording.md): Use record_llm_call, record_tool_call, and record_state to capture every event inside an AgentDbg traced run for inspection in the timeline.
- [Tracing Agent Runs with @trace and traced_run](https://refinehq.ai/docs/sdk/tracing.md): Learn how to start and manage traced runs in AgentDbg using the @trace decorator and the traced_run context manager for scripts and notebooks.