# Sovara

## Docs

- [Efficient annotation](https://docs.sovara-labs.com/annotations/annotate.md): Review surfaced runs and save useful labels.
- [Why annotate?](https://docs.sovara-labs.com/annotations/overview.md): Why annotation matters for improving agent behavior.
- [Recommendation algorithm](https://docs.sovara-labs.com/annotations/recommendation-algorithm.md): How Sovara decides what should be reviewed.
- [Annotations](https://docs.sovara-labs.com/cli/annotations.md): Queue, inspect, and label runs from the CLI.
- [Assistant skill](https://docs.sovara-labs.com/cli/assistant-skill.md): Install Sovara guidance for Codex and Claude Code.
- [Claude Code](https://docs.sovara-labs.com/cli/claude-code.md): Use the Sovara CLI with Claude Code.
- [Codex](https://docs.sovara-labs.com/cli/codex.md): Use the Sovara CLI with Codex.
- [Inspect runs](https://docs.sovara-labs.com/cli/inspect.md): Use probe, step-overview, and logs to understand a recorded run.
- [Install](https://docs.sovara-labs.com/cli/install.md): Install the Sovara CLI and verify that it can reach Sovara.
- [Lessons](https://docs.sovara-labs.com/cli/lessons.md): Read and manage Sovara lessons from the CLI.
- [Overview](https://docs.sovara-labs.com/cli/overview.md): Use the Sovara CLI to set up, report, inspect, rerun, annotate, and improve agent runs.
- [Report runs](https://docs.sovara-labs.com/cli/record.md): Run a command unchanged and print metadata for the SDK-created Sovara run.
- [Reference](https://docs.sovara-labs.com/cli/reference.md): CLI command reference for common Sovara workflows.
- [Replay keys](https://docs.sovara-labs.com/cli/replay-keys.md): Manage provider API keys used when rerunning recorded LLM steps.
- [Rerun LLM steps](https://docs.sovara-labs.com/cli/rerun-steps.md): Replay one recorded LLM step with edited input from the CLI.
- [Troubleshooting](https://docs.sovara-labs.com/cli/troubleshooting.md): Fix common Sovara CLI setup, reporting, inspection, and rerun issues.
- [Installation](https://docs.sovara-labs.com/get-started/installation.md): Install Sovara and prepare your local development environment.
- [Quickstart](https://docs.sovara-labs.com/get-started/quickstart.md): Record and inspect your first Sovara run.
- [Trace your agent](https://docs.sovara-labs.com/get-started/tracing-your-agent.md): Add structure and details to your logs.
- [Manual inspection](https://docs.sovara-labs.com/observability/manual-inspection.md): Inspect run steps, inputs, and outputs by hand.
- [Run chat](https://docs.sovara-labs.com/observability/run-chat.md): Use chat to query and reason about recorded runs.
- [Runs](https://docs.sovara-labs.com/observability/runs.md): Recorded executions in Sovara.
- [SDKs](https://docs.sovara-labs.com/sdks/index.md): Record agent runs from Python and TypeScript.
- [API reference](https://docs.sovara-labs.com/sdks/python/api-reference.md): The public Sovara Python SDK surface and exact client methods.
- [Python SDK quickstart](https://docs.sovara-labs.com/sdks/python/quickstart.md): Set up the Sovara Python SDK and record a project-bound agent run.
- [Troubleshooting](https://docs.sovara-labs.com/sdks/python/troubleshooting.md): Diagnose missing Python runs, steps, project assignment, and context.
- [Use the SDK](https://docs.sovara-labs.com/sdks/python/use-the-sdk.md): Understand project clients, runs, steps, subruns, metadata, and lessons.
- [API reference](https://docs.sovara-labs.com/sdks/typescript/api-reference.md): The public client, run, tracing, subrun, logging, and lesson APIs in @sovara/runner.
- [TypeScript SDK quickstart](https://docs.sovara-labs.com/sdks/typescript/quickstart.md): Set up the Sovara TypeScript SDK and record an agent run.
- [Troubleshooting](https://docs.sovara-labs.com/sdks/typescript/troubleshooting.md): Diagnose missing TypeScript runs, steps, project assignment, and noisy logs.
- [Use the SDK](https://docs.sovara-labs.com/sdks/typescript/use-the-sdk.md): Understand TypeScript project identity, runs, steps, subruns, metadata, and lessons.
- [Improve](https://docs.sovara-labs.com/sovara/improve.md): How Sovara turns observations into targeted improvements.
- [Observe](https://docs.sovara-labs.com/sovara/observe.md): How Sovara observes agent behavior during development.
- [Philosophy](https://docs.sovara-labs.com/sovara/philosophy.md): The principles behind Sovara’s approach to improving agents.
- [Surface errors](https://docs.sovara-labs.com/sovara/surface-errors.md): How Sovara helps expose failures and weak execution paths.
- [Sovara's auto-injection](https://docs.sovara-labs.com/sovaradb/auto-injection.md): How Sovara injects relevant lessons at runtime.
- [The context-problem of AI agents](https://docs.sovara-labs.com/sovaradb/context-problem-ai-agents.md): Why agent context needs durable memory and retrieval.
- [Creating and managing lessons](https://docs.sovara-labs.com/sovaradb/creating-managing-lessons.md): Create, organize, and maintain Sovara lessons.
- [What are lessons?](https://docs.sovara-labs.com/sovaradb/lessons.md): Define lessons and how they shape agent behavior.