> ## Documentation Index
> Fetch the complete documentation index at: https://doc.entelligence.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Agent Insights
> Measure coding-agent adoption, efficiency, spend, context health, and outcomes without treating raw activity as performance.
Agent Insights shows where AI spend goes, how coding agents are used, and what that activity produces. Use it to find workflow friction, guide enablement, and connect cost to engineering outcomes.
Agent Insights is in Beta. Available signals can vary by coding agent and workspace configuration.
## What you can answer
* **Where is spend going?** Compare repositories, projects, members, tools, and models.
* **Is agent usage healthy?** Inspect retries, interruptions, repeated edits, compaction, and context pressure.
* **What did the work produce?** Connect spend to features, fixes, tests, reviews, and exploration.
## Four views
| View | Use it for |
| ------------ | ------------------------------------------------------------------------------ |
| **Overview** | Adoption, efficiency, spend, context health, agent behavior, tools, and models |
| **Budget** | Project budgets, run rate, untracked spend, and weekly burn |
| **Outcomes** | Work categories, repositories, projects, members, teams, themes, and clusters |
| **Members** | Individual agent usage, efficiency, retries, skills, and activity |
## Read signals together
A retry spike may come from unclear tasks, missing context, tool failures, or a new workflow. Pair dashboard signals with the work being attempted and feedback from the team.
Do not use token volume, session count, efficiency, or adoption stage as a standalone measure of individual performance.
## Next step
Follow the [rollout quickstart](/agent-insights/setup), then learn how to interpret [metrics and data](/agent-insights/metrics).