How do I organize the troubleshooting materials when Hermes Agent reports an error?

When Hermes Agent reports an error and needs help, the most useful material is not a screenshot, but a hermes doctor, hermes dump, related logs, and minimal reproduction steps. This allows others to determine whether it's a configuration, dependency, model, platform, or tool issue.

It is recommended to collect these four types of information

1. hermes doctor Output: Used to check for dependencies, configurations, and missing environments.
2. hermes dump Output: Official commands provide a reproducible summary of settings, suitable for sending to supporters.
3. hermes logs relevant time periods: especially gateways, errors, platform access logs.
4. Minimal repro: What you typed, what you expected, what actually happened.

Pay attention to desensitization

Hermes logs do some secret redaction, but you should still check the API Key, Bot Token, Mailbox Password, Webhook secret, Server Address, and Private File Path before copying the material yourself. Be especially careful when posting to public communities.

How to make problems located faster

Don't just say "can't work". Explain the entrance: CLI, Open WebUI, Telegram, Discord, Webhook, MCP, or Docker. Then explain whether only one platform fails, or if all portals fail.

A good feedback format is: version and system, startup commands, related configuration snippets, full errors, steps to reproduce, fixes you have already tried. The more specific the material, the easier it is for a multi-component agent like Hermes to be quickly located.