use-ai
AI interaction primitive facade for RustUse.
Experimental
use-ai is experimental while the release line remains below 0.3.0.
Example
use ;
let prompt = new?;
let model = new?;
let tool = new?;
assert_eq!;
assert_eq!;
assert_eq!;
assert_eq!;
# Ok::
Scope
- Facade re-exports for focused AI interaction metadata crates.
- Prompt, message, context, role, model, provider, capability, tool-call, agent, reasoning, planning, RAG, memory, guardrail, and AI eval primitives.
Relationship to use-ml
use-ai models AI interaction primitives: prompts, messages, roles, context
windows, tool calls, agents, RAG, reasoning, memory, guardrails, AI model
interfaces, and AI-specific evaluation.
use-ml models machine-learning primitives: datasets, features, labels,
tensors, model artifacts, training, inference, evaluation, metrics, pipelines,
embeddings, experiments, and model documentation.
These sets are siblings. They should interoperate conceptually but avoid dependency cycles.
Non-goals
- Model calls, network calls, tool execution, retrieval, memory storage, guardrail enforcement, eval execution, or orchestration.
License
Licensed under either Apache-2.0 or MIT.