add HA and OC local apps

[merveenoyan]

adding logos to moon shortly

---

Note

Low Risk
Mostly additive snippet/predicate logic and test coverage for new local-app entries; low risk beyond potential UX gating/snippet formatting regressions.

Overview
Adds two new local app integrations, OpenClaw and Hermes (keyed as openclaw and hermes-agent), including copy/paste setup snippets for running them against a local OpenAI-compatible server.

Refactors the existing Pi integration by extracting a shared getLocalServerStep (MLX vs llama.cpp) and centralizing the “tool-calling local agent” eligibility check into isToolCallingLocalAgentModel, which is now shared by Pi/OpenClaw/Hermes. Expands tests to cover the new apps (GGUF + MLX cases) and asserts all three apps share the same displayOnModelPage predicate.

^{Reviewed by Cursor Bugbot for commit 66539fd. Bugbot is set up for automated code reviews on this repo. Configure here.}

[pcuenca]

I think these could belong in a setup phase, for symmetry with Pi (and uncomment the npm line).

[pcuenca]

Same, I'd make this a setup phase.

[gary149]

if not urgent would like to test it a bit before merging (when I'm back next week)

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 134a12f. Configure here.}

[merveenoyan]

@pcuenca moved to setup
@gary149 there's two ways of doing installation, one is onboarding interactively which doesn't require coding (it's better imho for inference providers but not for local), the other one is non-interactive for which you get llama-server up and setup like this or add the endpoint to your config