A professional v3 observability toolkit for OpenClaw agents featuring automatic tracing, cost analysis, and system health monitoring.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install langfuse-observability
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install langfuse-observability using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Langfuse Observability?
Langfuse Observability is designed as the central nervous system for monitoring and debugging AI agents within the Openclaw Skills ecosystem. It provides a robust v3 toolkit that allows developers to implement deep observability for LLM calls, API requests, and tool executions. By integrating this skill, users gain complete visibility into their agentic workflows, moving from black-box operations to fully transparent and auditable processes.
The skill excels at synthesizing complex data points into actionable insights. It captures everything from high-level session groupings to granular cost tracking per model, ensuring that developers can maintain both performance and budget efficiency. Whether you are running locally or in a distributed cloud environment, this skill provides the telemetry needed for production-grade reliability.
Langfuse Observability Use Cases
- Debugging complex multi-step agent workflows to identify and resolve logic failures or bottlenecks.
- Monitoring LLM API costs in real-time to manage budget constraints per session or per model.
- Auditing agent interactions for compliance and quality assurance using detailed trace logs.
- Automated health monitoring of agentic systems with scheduled daily status reports.
How Langfuse Observability Works
- The skill is integrated into the workspace by pointing the Python system path to the langfuse-observability scripts directory.
- Developers wrap their agent functions with decorators like @traced or use specific methods such as trace_llm and trace_tool to capture execution data.
- The Langfuse Python SDK v3.14.1 automatically transmits telemetry data to the designated Langfuse host.
- Traces are organized in the dashboard where they are grouped by session and analyzed for token usage and associated costs.
- Admin and cron scripts enable CLI-based queries and automated reporting to external channels like Telegram.
Langfuse Observability Setup
To get started with these Openclaw Skills, you need to configure your environment to recognize the observability scripts. Use the following setup in your agent scripts:
# Add the skill directory to your Python path
import sys, os
sys.path.insert(0, os.path.expanduser("~/.openclaw/workspace/skills/langfuse-observability/scripts"))
# Import the tracing toolkit
from langfuse_hub import traced, trace_llm, trace_api, trace_tool, trace_event, flush
Ensure your Langfuse instance is running at http://langfuse-web:3000.
Langfuse Observability Data Schema & Taxonomy
This skill utilizes the Langfuse v3 schema to organize observability data into a hierarchical structure:
| Component | Description |
|---|---|
| Traces | The top-level execution context for an entire agent run. |
| Spans | Represent individual units of work or tool executions within a trace. |
| Generations | Specific records of LLM prompts, completions, and token counts for cost calculation. |
| Scores | Qualitative or quantitative evaluation metrics attached to specific traces or spans. |
Langfuse Observability Advanced Features
- Automatic cost tracking per model version to monitor infrastructure spending.
- Session grouping capabilities to link disparate agent events into a single user journey.
- Evaluation scoring to help benchmark agent performance over time.
- CLI-based administrative tools for querying session health and audit trails directly from the terminal.
- Integrated cron jobs for automated daily status reports delivered to team communication channels.
SKILL.md
Loading
Related Openclaw Skills
Inventory Source
A comprehensive automation tool for managing dropship suppliers, product feeds, and order routing through a unified CLI.
IFTTT
A zero-dependency Python skill for triggering IFTTT webhooks and managing applets to connect AI agents with 800+ external services.
Hunter.io Email & Lead Management
A comprehensive Hunter.io integration for AI agents to find, verify, and manage professional email leads via CLI.
Healthie
Manage clinical workflows including patients, appointments, and documentation through a streamlined GraphQL API interface.
Lead Gen Pipeline
An AI-powered automation skill for scoring leads and generating context-aware sales follow-ups.
Lemon Squeezy
A comprehensive CLI tool for managing Lemon Squeezy digital products, orders, and subscriptions within Openclaw Skills.
Featured*
