What model are you using with Metron?

[mdabydeen]

Metron is early and the model boundary matters. If you try the v0.1.0 release, I would like to learn which local model and hardware you used, which tools worked, and where the interaction broke down.

Please include:

• model name and provider
• operating system and architecture
• whether metron --doctor passed
• the task you tried
• any tool-call, context, or patch-approval issue you saw

This is a request for concrete compatibility feedback, not a claim that every model or setup is supported. The release and installation path are documented here: https://github.com/mdabydeen/metron/releases/tag/v0.1.0

[mdabydeen]

Baseline from the maintainer's workstation (27 September 2026): gemma4:12b-mlx on macOS arm64 with local Ollama. metron --doctor returned doctor: ready after Universal Ctags was placed first on PATH. A no-write one-shot list_files check for the repository-root README.md executed successfully and returned yes.

This is one setup and task observation. It does not establish general compatibility, performance, support, or production readiness. If you try the release, please include the model/provider, OS and architecture, metron --doctor result, task, and any tool-call, context, or patch-approval issue.

[mdabydeen]

Additional first-run observation (28 September 2026): I ran the published Darwin arm64 v0.1.0 archive in a clean temporary directory. metron --version reported v0.1.0; metron --doctor correctly reported that the directory was not a Git repository and that the built-in placeholder model qwen2.5-coder:32b was not installed. Running inside the target repository with a configured tool-capable Ollama model is therefore part of the setup. I clarified this path in the main README; the release binary itself was not rebuilt. This is setup evidence, not a general compatibility or performance claim.