Add hermes-claude-code-local to Coding Agents section

[KaiFelixBennett]

What this adds

hermes-claude-code-local - run Claude Code and Hermes Agent locally on llama.cpp via LiteLLM — zero API costs, no rate limits, full privacy.

Why it fits

• Coding agent setup (Claude Code + Hermes Agent) running fully locally
• Uses llama.cpp as the inference backend (fits the 'local LLM' theme)
• LiteLLM bridges the Anthropic API format to any local OpenAI-compatible endpoint
• Real-world validated: 4-hour autonomous coding session, 7.2M tokens, \ cost (vs \ on Claude Opus 4.7)

Category

Fits in Coding Agents — it's a complete local coding agent stack rather than a standalone model or inference engine.

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.