Technical writing from the Chrysent team: engineering notes, design rationale, and points of view on governing enterprise AI. Company announcements live under News; product updates live on each product page.
Security
AI gateway security lessons from recent LiteLLM, OpenCode, Railway, and GitLab incidents—and how Chrysent keeps provider keys out of client applications while governing usage inside your environment.
Engineering
Use rust-sitter-mcp to help coding agents find useful Rust module boundaries, plan supported moves, and review the dependencies before changing files.
Engineering
How to compare local coding models on an M2 Pro Mac, measure cache and MTP behavior, and connect Pi through llama.cpp’s native router.
Case study
See how human-approved security requirements, specialist AI agents, blocking review gates, and live validation were used to deliver sandboxctl.
Engineering
How must-fix verdicts, specialist reviewers, and focused re-review turn coding-agent output into a governed software-delivery process.
Research
A layered model for validating AI-generated software through source authority, traceability, tests, live acceptance, exact cleanup, and explicit limitations.
Engineering
See how AI agent harnesses control context, token costs, latency, and software quality—and how Chrysent is applying the harness effect in OPI Forge.
Product
Start executing with what you have, then keep evolving and adapting with OPI Forge.