Awesome AI AgentsAgent Monitoring Tools

RyjoxTechnologies/Octopoda-OS

⭐ 491 Python repository created 2026-04-02

Octopoda is an MIT-licensed Python library that sits underneath an existing agent and supplies four things that agent frameworks usually leave out: memory that survives restarts, detection of stuck loops, an audit trail of decisions, and a dashboard for watching a fleet of agents. Integration is deliberately shallow. After a pip install, a single init call auto-detects the framework in use, captures what matters from each turn, distills it into memories and injects relevant recall into later model calls, without changes to the agent logic. A wrapper command can also instrument an unmodified agent script at launch. The library works with plain Python as well as LangChain, CrewAI, AutoGen, the OpenAI Agents SDK and MCP, and it also exposes a direct SDK where an agent runtime object remembers and recalls named values that persist across process restarts. Loop detection runs automatically on every write and recognises retry, oscillation, ping-pong, reflection and recall-write patterns, surfacing the exact calls responsible; intervention such as auto-pause or a spend cap is opt-in through a circuit-breaker configuration, so the policy stays with the operator. Audit events written through the versioned audit endpoint are hash-chained per agent, linking each event to the previous one so integrity can be verified with one call. The dashboard shows agent health, operation volume, per-agent scores, an anomaly stream and the loops that were caught, and the same interface runs locally against local data with no account or against a hosted cloud backend when an environment variable is set. Octopoda also exposes an MCP server. It is aimed at developers who have shipped an agent into production and then could not explain what it did, or watched it burn tokens in a retry loop unnoticed.

https://github.com/RyjoxTechnologies/Octopoda-OS

agent-frameworkai-agentsai-memoryautogencrash-recoverycrewaideveloper-toolsknowledge-graphlangchainllm-agentsloop-detectionmcpmcp-servermemory-enginemulti-agentobservabilityopenaipersistent-memorypythonsemantic-search

Also in Agent Monitoring Tools

raga-ai-hub/RagaAI-Catalyst

RagaAI Catalyst is a Python SDK platform for observability, monitoring, and evaluation of AI agents and LLM projects, featuring tracing, debugging, advanced analytics, and project management tools.

superset-sh/superset

Superset is a desktop application (macOS) designed to orchestrate and manage multiple CLI-based AI coding agents in parallel using isolated Git worktrees, featuring built-in diff viewing and IDE in...

MervinPraison/PraisonAI

PraisonAI is a production-ready Multi AI Agents framework that enables the creation and management of intelligent AI agents to automate and solve complex problems through collaboration and customization.

apache/hertzbeat

Apache HertzBeat is an AI-powered open source real-time observability system that unifies metrics and logs collection, alerting, and notification with high performance and customizable monitoring capabilities.

junhoyeo/tokscale

Tokscale is a CLI utility and dashboard for tracking and visualizing token usage and costs across various AI coding agents.

matt1398/claude-devtools

Desktop and Docker application that reads local Claude Code session logs and reconstructs tool calls, thinking steps, subagent trees and per-turn token attribution in a visual interface.

eumemic/ai-legion

AI Legion is a platform for autonomous agents powered by large language models that collaborate to accomplish tasks with minimal human intervention.

chiphuyen/sniffly

Sniffly is a local analytics dashboard that parses Claude Code logs to show usage statistics, error breakdowns and full message history, with optional shareable dashboards.