# Captar ## Docs - [OpenAI Wrapping](https://docs.captar.aurat.ai/core-concepts/openai-wrapping.md): How wrapOpenAI keeps your existing client while attaching policy, estimation, spend reconciliation, and spans. - [Sessions and Budgets](https://docs.captar.aurat.ai/core-concepts/sessions-and-budgets.md): How Captar sessions hold budget state, metadata, policy, and the root trace span. - [Tool Guardrails](https://docs.captar.aurat.ai/core-concepts/tool-guardrails.md): Track tool spans, apply tool policy, and keep external actions inside the trace. - [Traces and Export](https://docs.captar.aurat.ai/core-concepts/traces-and-export.md): Span-first traces, exported events, and how Captar turns strong traces into reusable dataset rows. - [Installation](https://docs.captar.aurat.ai/getting-started/installation.md): Add the Captar SDK, prepare environment variables, and set up the base project shape. - [Overview](https://docs.captar.aurat.ai/getting-started/overview.md): The shortest path to a working Captar integration for a TypeScript app using OpenAI. - [Quickstart](https://docs.captar.aurat.ai/getting-started/quickstart.md): Start a session, wrap OpenAI, track a tool call, and flush exported events. - [Captar](https://docs.captar.aurat.ai/index.md): Runtime control for OpenAI apps, with traces, budgets, and manual review built into the docs path. - [Datasets](https://docs.captar.aurat.ai/platform/datasets.md): Build append-only datasets from trace exports or file imports, then use them for manual review. - [Manual Evals](https://docs.captar.aurat.ai/platform/manual-evals.md): Launch reviewer runs from datasets and score rows with pass or fail plus rubric criteria. - [Trace Inspection](https://docs.captar.aurat.ai/platform/traces.md): Use the platform trace page to inspect span trees, payloads, spend, and violations. - [Configuration and Environment](https://docs.captar.aurat.ai/reference/configuration.md): Exporter, policy, pricing, and environment setup for the Captar SDK. - [Events and Types](https://docs.captar.aurat.ai/reference/events-and-types.md): The main event, policy, trace, dataset, and manual eval types that shape v1 behavior. - [SDK API](https://docs.captar.aurat.ai/reference/sdk-api.md): Curated reference for the main Captar SDK entrypoints used in v1. ## OpenAPI Specs - [openapi](https://docs.captar.aurat.ai/api-reference/openapi.json)