# Codenotary releases AgentMon Live open source tool for Claude Code monitoring

> **Key takeaway:** AgentMon Live is an Apache 2.0 licensed tool designed to monitor AI coding agents, verify task completion, and reduce token waste.

**Published:** 2026-11-11T21:44:09+00:00 
**Source:** FOSS Force 
**Category:** distro-news 
**Canonical URL:** https://fosswire.org/news/codenotary-releases-agentmon-live-open-source-tool-for-claud.html 

## Summary
Codenotary has announced the release of AgentMon Live, an open-source monitoring tool specifically designed to assist developers using Claude Code. The software aims to mitigate the common issues associated with AI coding agents, such as excessive token consumption, unverified task completion, and the high cognitive load required for human supervision.

Licensed under Apache 2.0, the tool provides a real-time dashboard and a verification layer known as 'Done Gate.' It is designed to run locally and is compatible with Claude Code version 2.1.308 and later. The project is intended to help developers identify when an agent has failed tests or left tasks unfinished before reporting a job as complete.

## Technical capabilities of AgentMon Live
- **Done Gate Verification:** This feature examines recorded session activity to detect unresolved failures, such as failed or outdated tests and incomplete subagent tasks, prompting the agent to rectify the work.
- **Real-time Monitoring Dashboard:** Provides a single interface to track active tools, subagent status, file modifications, pending approvals, and Model Context Protocol (MCP) server usage.
- **Token and Efficiency Tracking:** Monitors token consumption, context growth, and cache efficiency while identifying inefficiencies like repeated file reads or excessively large tool outputs.
- **Local Execution and Extensibility:** The dashboard and verification checks run locally on the developer's machine, and the tool supports an optional second-model advisor to assess if agent responses meet original requests.
- **Benchmarking Integration:** Includes a built-in command that allows developers to compare performance and efficiency metrics between sessions with and without the tool active.

## Why it matters
As AI agents become more integrated into software engineering workflows, tools like AgentMon Live provide necessary guardrails for maintaining code quality and controlling API costs. By offering an open-source monitoring layer, it allows developers to audit agentic behavior and ensure that automated code generation adheres to verification standards.