rustic-ai 0.2.0

A Rust-native agent framework with tool calling, streaming, and multi-provider support for OpenAI, Anthropic, Gemini, and Grok
Documentation
1
2
3
4
5
6
7
8
9
10
- Rust safety and reliability:
  - Avoid `unsafe` in production; if required, isolate and document invariants/assumptions.
  - Validate/sanitize external inputs (headers, query/body); use parameterized queries.
  - Prefer `Option`/`Result` and `?` for propagation; avoid `unwrap`/`expect` on external data.
    - Tests may use `unwrap`/`expect` for setup and assertions.
  - Avoid blocking calls in async contexts; use async I/O or `spawn_blocking` for CPU-bound work.
  - Enable overflow checks in release builds in `Cargo.toml`.
- Mock external deps in tests (Twilio, OpenAI, etc.)
- #![allow(dead_code)] is not allowed. No compiler warnings, no clippy warnings.
- Never commit broken tests