Open-source debugging and run-inspection tool for AI agent workflows, prompts, tool calls, and runtime state.
License
Not available
Stars
23
Features
- Agent execution inspection
- Prompt and context visibility
- Tool-call debugging support
- Useful for local development and experiments
- Open-source integration fit
- Reduced black-box debugging cost
Use Cases
- Agent workflow debugging
- Local prototype experiments
- Prompt-chain troubleshooting
- Tool-call behavior analysis
- Internal agent platform integration
- AI application quality review
FAQ
AgentPeek is an open-source project for AI agent development and debugging. It helps developers inspect execution flow, prompt context, tool calls, and intermediate state, making it useful for local experiments, framework integration, and internal agent workflow iteration. Repository: https://github.com/TranHuuHoang/agentpeek.
Common real-world fits include: Agent workflow debugging, Local prototype experiments, Prompt-chain troubleshooting.
