Expand description
Multi-provider LLM abstraction for ResQ developer tools.
Supports Anthropic, OpenAI, and Google Gemini APIs with a unified
complete() interface and cascading config resolution.
Structs§
- AiConfig
- AI configuration.
Enums§
- Provider
- Supported LLM providers.
Functions§
- complete
- Send a completion request to the configured provider.
- estimate_
tokens - Estimate token count using the chars/4 heuristic.
- load_
config - Load config with cascade: env vars -> ~/.resq/ai.toml -> .resq/ai.toml.
- truncate_
to_ budget - Truncate text to fit within a token budget. Cuts at line boundaries to avoid broken diff hunks.