AgentsView discovers coding agent sessions on your machine, syncs them into a local SQLite database, and serves a web UI with token usage dashboards, cost tracking, and activity analytics. It supports over 20 agents and serves as a faster replacement for ccusage, with features like live session streaming, activity heatmaps, and per-model cost breakdowns. All data stays local with no cloud dependency.
AgentsView is a local-first analytics platform that gives you visibility into your coding agent sessions. It automatically discovers sessions from Claude Code, Codex, and over 20 other agent tools on your machine, syncs them into a local SQLite database, and serves a web dashboard with usage analytics, cost tracking, and activity insights.
The platform provides real-time session monitoring with live updates via server-sent events, so you can watch active agent sessions as they run. The analytics layer tracks token usage per model and per session, calculates costs, generates activity heatmaps showing when you use agents most, and breaks down tool usage patterns. Everything runs locally — your session data never leaves your machine.
AgentsView is built for developers who use coding agents daily and want to understand their usage patterns. Track how much you spend on API calls across different models, identify which projects consume the most tokens, monitor active sessions in real time, and compare agent performance across tools. Teams can use it to manage shared API budgets and identify optimization opportunities.
AgentsView reads session data from local agent directories. It works best with agents that store structured session logs (Claude Code, Codex, Aider, etc.). Some agents may require additional configuration for session discovery. The tool is focused on analytics and monitoring — it does not modify or control agent behavior.
AgentsView is for developers and teams who use AI coding agents regularly and want visibility into usage, costs, and patterns. Power users tracking API spend, team leads managing shared budgets, and anyone curious about their agent usage will find it valuable.
Tool-output sandboxing layer that cuts agent context spend 60-98% on multi-step coding loops by isolating intermediate command output.
AI-powered observability platform that provides intelligent monitoring, anomaly detection, and automated insights across your entire tech stack.
AI-powered incident management platform that reduces alert noise, automates triage, and accelerates incident resolution.