Skip to content
TriedPrompts

More in Team workflows & setup.

Fix a self-hosted agent that lost its models: Muse vs Hermes

PartialEvidence: Described by posterPosted Last checked

Tun Win Lat asked Muse to restore missing Qwen models; it wrongly said it was fixed three times, and Hermes fixed it in one go.

How to use this

  1. Tap Copy prompt.
  2. Open the tool the case used (or a similar one) and paste the prompt; results may differ from the original.
  3. Paste it into the message box.
  4. Send it, then read the answer before acting on it.
New to this? What's a prompt?

A prompt is the message you send to the AI tool. Copy the one on this page, put your own details where it asks, and read the answer before you act on it.

Prompt

Recreated prompt

The source didn't share its exact prompt, so we wrote one from what it describes. The result shown is the source's, not ours.

What the prompt labels mean

The person didn't share their exact words, so we wrote this from what they described. We haven't tested it, and they didn't use this exact version.

My OpenClaw setup has lost its Qwen models. Find out why and fix it, and confirm the models are working again before you tell me it's done.

Try this in:

ChatGPT Gemini Grok Claude

Sends the Recreated prompt shown above. ChatGPT and Grok open with it filled in, and ChatGPT may send it straight away. Gemini and Claude open empty, so we copy the prompt for you to paste. Every button also copies it.

The source used Muse and Hermes. These are general chat tools, so results may differ.

Opens in a new tab. We don't see what you send.

These links are only a shortcut. We're not affiliated with OpenAI, Google, Anthropic or xAI, and your results may not match the source.

Optional stronger version

Our rewrite with extra checks. The poster didn't test it.

Before you start, ask me 1–3 short questions about the details below and wait for my answers. Once I've answered, confirm the plan in one line and run the whole task start to finish.

Goal: Find and fix why my self-hosted agent lost its models, and prove the fix works before calling it done.

Questions to ask me first:
- What changed right before the models disappeared, such as an update or config edit?
- Where do the config files and logs live?
- Can you run commands on the server, or should I run them?

Inputs (fill in or confirm):
- Agent and version: [FOR EXAMPLE OPENCLAW VERSION]
- Missing models: [MODEL NAMES]
- Config and log locations: [PATHS]
- Recent changes: [UPDATES OR EDITS]

Rules:
- Read the logs and config before changing anything.
- Back up any file before you edit it.
- Make one change at a time and test after each.
- Never say it's fixed until a test call to each model succeeds.

Checkpoint: Tell me the likely cause and your planned change before editing anything.

If you get stuck: If the first fix doesn't work, say so plainly and show what you saw, instead of trying again quietly.

Verify before you finish: Run a test prompt through each model and paste the output.

Output: The cause, the change made, the backup location and the test results.

Done when: Every model answers a test prompt, and I can see the proof.

AI tools

MuseHermes

Outcome

Tun Win Lat runs six AI agents day to day, including a self-hosted OpenClaw instance for personal operations. When that OpenClaw setup lost its Qwen models, Tun asked Muse, the everyday assistant, to fix it. Over three attempts, Muse said each time that the problem was fixed, and it wasn't. Tun took the same problem to Nox, a self-hosted Hermes agent from Nous Research that handles dev work and was already running a Qwen model. Nox diagnosed the actual issue and fixed it in one go. Tun still uses Muse as the daily driver but notes that it doesn't always know where its limits are.

Evidence: Described by posterThe person describes what happened in words, without screenshots or video.

Known limitations

PartialThe source says it only partly worked, or had a significant problem a person had to deal with.

What went wrong: Muse claimed a fix three times without fixing anything; a second agent, Hermes, solved it in one attempt.

Source

Posted
by Tun Win Lat (blog)
Last checked
Method:
How we verify

https://www.tun.lat/post/six-ai-agents-no-single-winner

Context: The exact request isn't quoted, so the prompt here is recreated. This is one story from a wider post comparing six agents Tun uses every day.

More in Team workflows & setup

Similar cases