Add integrated LLM chat mode to Walker

[Rishabh-Sarang]

Describe the feature Add an optional built-in LLM chat mode inside Walker. Users could type natural-language questions (e.g., “What’s the weather?”, “Explain recursion”, “Summarize this text”) directly into the launcher and get answers from an LLM.

This feature would make Walker a lightweight, fast-access AI assistant — no need to open a separate browser or app.

The LLM backend should be configurable, supporting:

• Local models (like Ollama, llama.cpp, etc.)
• Remote APIs (e.g., OpenAI, Anthropic, etc.)

Alternatives

• Continue using external chat apps (browser-based or terminal-based LLM clients).
• Use a plugin system, if available, to implement chat via an external extension.

Describe the behaviour

• Activation:

1. User opens Walker as usual.
2. Presses a hotkey or enters a trigger (e.g., /ai or ?) to enter Chat Mode.

• Input:

1. User types a natural-language prompt, e.g.:
2. “/ai What is the fastest sorting algorithm?”
3. “Explain the difference between TCP and UDP.”

• Processing:

1. Walker sends the prompt to the configured LLM backend.
2. The backend could be local (Ollama, LM Studio) or API-based (OpenAI, etc.), depending on user settings.

• Output:

1. The model’s response is displayed directly in Walker — either inline or in a small popup/chat pane.
2. User can press Shift+Enter (or similar) to continue the conversation contextually.
3. Optionally allow copying the response with one key.

• Settings:
• Users can configure:

1. Model backend and endpoint
2. API key (if needed)
3. Response length / temperature
4. Whether chat history is saved or cleared on close

[stopdesign]

Or maybe no? Maybe an app launcher shouldn't be an AI assistant?

[Rishabh-Sarang]

Or maybe no? Maybe an app launcher shouldn't be an AI assistant?

walker used to have that feature