Add multi-LLM provider support and position agent system for cross-modal geophysics

[Copilot]

Positions the agent system as "Automatic cross-modal geophysics agent for subsurface hydrology" supporting multiple geophysical data types (ERT, seismic, etc.) and multiple LLM API providers (OpenAI GPT, Google Gemini, Anthropic Claude) for publication.

Core Changes

• BaseAgent: Added llm_provider parameter with provider-specific API key detection, default models, and query implementations

  • OpenAI: OPENAI_API_KEY, gpt-4
  • Gemini: GEMINI_API_KEY, gemini-pro
  • Claude: ANTHROPIC_API_KEY, claude-3-opus-20240229

• AgentCoordinator: Added llm_provider parameter with automatic provider-specific key resolution

• Dependencies: Added google-generativeai>=0.3.0 and anthropic>=0.18.0 to agents extras

Documentation Updates

• Updated all references from "ERT-only" to "cross-modal geophysics" supporting multiple data types
• Positioned system for paper: "Automatic cross-modal geophysics agent for subsurface hydrology"
• Updated README, examples, and package descriptions to reflect broader capabilities

Usage

# OpenAI (default)
coordinator = AgentCoordinator(llm_provider='openai')

# Google Gemini
coordinator = AgentCoordinator(llm_provider='gemini')

# Anthropic Claude
coordinator = AgentCoordinator(llm_provider='claude')

# Custom model
coordinator = AgentCoordinator(
    llm_provider='gemini',
    model='gemini-1.5-pro'
)

Backward compatible: existing code defaults to OpenAI provider.

Original prompt

in the new agent part. not only ERT, in the future, it can get other geophysical data into hydrologic information. I want to use this agent to publish a paper like "Automatic cross‑modal geophysics agent for subsurface hydrology." so update the description. also not only GPT API. we can also add others like gemimi and Claude API.

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.