Mog, a programming language for AI agents
Sandbox-aware language: agents can enforce permissions even on Mog-written bash commands.
Jo Secure Programming Language
Compiler-enforced capabilities catch prompt injection before AI-generated code ever runs.
Developers building AI agent platforms, security-conscious teams running untrusted code
Pony · WASI · Rust
Sandbox-aware language: agents can enforce permissions even on Mog-written bash commands.
First language designed for LLMs to modify safely with capability permissions.
Built-in reactive runtime and dual JS/WASM targets challenge React and TypeScript dominance.
Endianness in the type system is clever, but another language in a saturated field.
Static scanner catches prompt injections in code before runtime, unlike runtime guards.
Design-by-contract baked into grammar for a no-GC systems language.