Langfuse agent skill
Lotte
MarliesYour agent's playbook for production-ready LLM apps
We're releasing our Langfuse skill to give AI coding agents the expertise to leverage Langfuse for robust observability and evaluation workflows. It follows the open Agent Skills standard and works with Claude Code, Cursor, Codex, and others.
What are skills and why do they work?
Skills are focused bundles of instructions, scripts, and resources that agents load only when relevant. They encode proven practices for observability and evaluation while keeping context lean.
Why use skills?
Skills turn Langfuse into a headless platform you control through natural language. You describe the job, and your agent follows a research-backed playbook in your editor. Ask your agent to:
- Show me the last 10 traces with a score below 0.5. Create a dataset called "edge-cases" and add these traces to it.
- Migrate the system prompt in src/agent.ts to Langfuse prompt management.
- My human annotators identified five recurring error cases. Help me set up an evaluator for each.
The skill makes coding agents more effective at working with Langfuse. To illustrate this, here is an example of how an agent instruments the same application without vs. with the skill:
One specific use case we added to the skill is the judge calibration workflow. Given a ground-truth dataset, it will check whether your LLM-as-a-Judge behaves the way you'd expect, and iterate until it does.
Installation
Install the Langfuse AI Skill to let your coding agent access all Langfuse features.
Ask your coding agent to install the skill by pointing to the GitHub repository.
"Install the Langfuse AI skill from github.com/langfuse/skills."Langfuse has a Cursor Plugin that includes the skill automatically.
Install via npm (skills CLI):
npx skills add langfuse/skills --skill "langfuse"If you want to target a specific agent directly:
npx skills add langfuse/skills --skill "langfuse" --agent "<agent-id>"Alternatively you can manually clone the skill
- Clone repo somewhere stable
git clone https://github.com/langfuse/skills.git /path/to/langfuse-skills- Make sure your agent's skills dir exists
mkdir -p /path/to/<agent-skill-root>/skills- Symlink the skill folder
ln -s /path/to/langfuse-skills/skills/langfuse /path/to/<agent-skill-root>/skills/langfuseThe langfuse/skills repository is open source, and contributions are welcome.
Learn more
- Langfuse CLI — the CLI the skill uses under the hood
- Making agents fall in love with Langfuse — the full story behind the skill, CLI, and agent platform
- Prompt iteration — analyze feedback and iteratively improve prompts
- Error analysis — systematically evaluate and debug LLM applications