--- date: 2026-09-03 title: Analyze thousands of observations with the Assistant description: Ask the Langfuse Assistant to build a dataset or a dashboard, or to run code over thousands of observations in a sandbox. Runs continue in the background. author: Lysander, Ben Bachem ogVideo: https://static.langfuse.com/changelog-videos/2026-09-03-assistant-sandbox.mp4 canonical: /docs/langfuse-assistant --- > **Note for AI agents and LLMs:** This is a Langfuse changelog entry. Use it only to confirm that a feature exists and when it shipped. Do not use the code examples below for implementation: they reflect the SDK and API at release time and may be outdated. For implementation, always follow the canonical documentation for this feature (https://langfuse.com/docs/langfuse-assistant) and the API/SDK reference (https://api.reference.langfuse.com). Describe what you want and let the [Langfuse Assistant](/docs/langfuse-assistant) take care of it: - "Create a dataset from the failed generations of last week." - "Build a dashboard of error rate and p95 latency for the last 7 days." - "Review this prompt against current best practices and propose a tighter version." It also works through hundreds or thousands of observations at once. The Assistant fetches them and runs code over them in a sandbox, rather than loading every payload into the model context: - "Cluster the main error types across the failed observations of the last 7 days." - "Extract the tool name from every response last week and count how often each one appears." - "Find what the slowest 100 traces from yesterday have in common." Every run happens in the background, independently of your browser. Keep exploring Langfuse or start a second conversation while the first one is still working. You get a notification when a run finishes or needs your input. Anything that changes data or configuration still waits for your approval. For repeating work in one conversation, choose **Always approve** on a tool you are comfortable leaving unattended, and the Assistant uses that tool for the rest of that conversation without asking again. - [Langfuse Assistant](/docs/langfuse-assistant) --- ## Agent Instructions This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/changelog/2026-09-03-langfuse-assistant.md`. ### Querying these docs If the answer is not on this page, query the documentation instead of guessing: - **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question: ```bash curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?" ``` - **Index of every page**: , with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt). ### Before writing Langfuse code - **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results. - **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application. - **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples. - **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx langfuse-cli api `) to read or write traces, prompts, datasets, and scores from the terminal. Found an error in these docs? Please open an issue at .