Build Tool-Calling Parsers in Rust with AI Agents
A new guide explains how to process and parse fragmented or incomplete tool calls from AI models in Rust, converting complex outputs into application-ready actions.
Developer intelligence
What changed, why it matters, and what you can do now.
Your stack
Add your stack to personalize.
A new guide explains how to process and parse fragmented or incomplete tool calls from AI models in Rust, converting complex outputs into application-ready actions.
Large-scale, AI-driven Rust rewrites of the curl codebase have not kept pace with ongoing bugfixes and security patches in the official C project, raising practical concerns for developers about using or adopting these clones.
The LiteLLM v1.109.0-dev.2 release introduces new provider support, enhanced quota policies, auditing, major bug fixes, Docker image signature verification with cosign, and multiple breaking changes relevant to all LiteLLM Docker deployments and API users.
LiteLLM v1.105.0 introduces cryptographic Docker image signing, enhanced provider and integration support (including xAI and Sail), budget alerting, and new safety mechanisms for proxy rate limiting and configuration cleanup.
Parseable, a new open-source observability platform, now works with Datasette's recent OpenTelemetry support, allowing developers to feed and visualize trace data from Datasette into Parseable.
LiteLLM's latest release brings Docker image signing with cosign, breaking changes to proxy auth requiring a strong master key, new password security features, and support for Seedance video generation via FAL queue API.
The latest LiteLLM release adds Docker image signature verification with cosign, introduces a new OpenAI-compatible provider (Cortecs), enhances tracing with ClickHouse storage and OTLP ingestion, and refactors the internal Lens tracing engine with breaking API changes.
sglang v0.5.21 introduces support for top open LLMs and diffusion models, major API updates for low-latency scoring/classification, performance boosts for popular models, and a new Rust prefix cache.