---
icon: zy/agent
---
# Agent Monitoring
---
Agent Monitoring correlates Agent/LLM requests with the entire application trace, tracking the complete flow of each conversation and precisely measuring the token consumption of each generation task.
When using Agent Monitoring, you can:
- View the complete trace of a single request: clearly see the entire process from user query reception, processing (e.g., database query), to LLM model invocation and response generation
- Analyze performance bottlenecks: precisely measure the duration of each phase (e.g., model invocation, data retrieval) and detect latency in a timely manner
- Correlate upstream and downstream services: correlate Agent/LLM requests with related application and infrastructure metrics for comprehensive root cause analysis
## Core Capabilities
The core of Agent Monitoring is establishing a quantifiable link between input (Prompt), output (Completion), and system behavior. Its capabilities are reflected in three dimensions:
**1. Full Trace**
Within the Agent/LLM invocation framework, use traces and spans to precisely track the complete request flow and pinpoint latency bottlenecks.
**2. Quality Output Evaluation**
Internally, a rule engine and AI evaluation automatically optimize output content.
**3. Cost Measurement**
Automatically collect and correlate the token consumption (input/output breakdown), model type, and invocation parameters for each request, providing cost allocation capabilities based on multiple business dimensions.
## Getting Started
### Application List
Enter the **Application List** to create and manage monitoring applications. You can choose to create a new [**Agent Monitoring Application**](./agent-apps.md) or [**LLM Monitoring Application**](./llm-apps.md). Agent Monitoring Applications support Codex, Claude Code, OpenClaw, Hermes, Qoder, WorkBuddy, OpenCode, and AgentScope. AgentScope is manually instrumented via OpenTelemetry; other types use the `obs-agent-connector` installer for instrumentation. LLM Monitoring Applications support ingestion via Langfuse. After completing the parameter configuration guided on the page, you can start collecting data.
The summary page allows you to customize the Agent types displayed and adjust their order via the "Display Items" option.
### Explorer
After data ingestion, you can search and filter data by Session or Trace dimension in the [**Explorer**](./explorer.md):
- **User List**: Aggregates user usage by `user_id`, displaying Agent type, instance, session, request count, token or credit consumption, and last active time
- **Session List**: Displays Session ID, input/output tokens, number of risk events, and alert severity. Drill down to the detail page to view the trace waterfall chart, model/Skill/Tool invocation ratios, invocation details, and risk events
- **Trace List**: Displays Trace ID, associated session, duration, tokens, status, number of risky operations, and alert severity. Drill down to view span details, risky operations, tool invocation records, and Skill invocation records
### Analysis Dashboard
Use the [**Analysis Dashboard**](./analysis-dashboard.md) to view an overall picture of application health in chart form. You can filter by application type. The content varies by application type:
- **LLM Monitoring Application**: Displays overview metrics such as request count, span count, request error rate, total token consumption, and average response latency, as well as request trend charts, token consumption trend charts, request share by model, and token usage ranking by model.
- **Agent Monitoring Application**: Displayed in four modules: **Run Overview**, **Model & Token**, **Tool Module**, and **Skill Module**.
- Run Overview: Total operations, total Tool, LLM, and Skill invocations, total tokens, and distribution of model/Tool/Skill invocations
- Model & Token: LLM invocation trend, input/output token trend, token trend by model, and LLM invocation duration percentiles
- Tool Module: Active tool count, maximum duration, invocation trend, average duration, and anomaly Top 10
- Skill Module: Active skill count, maximum duration, invocation trend, average duration, and anomaly Top 10