capability-security

Astrid is a portable, capability-secure operating system for composable software.

Jacquard is a small programming language designed for a regime in which most code is written by machine-learning models and reviewed by people.

An actor-based systems language that compiles to readable C. Native, no VM, no garbage collector.

Sandboxed plugin VM with typed capabilities, deterministic replay, and time-travel debugging — written in Rust.

Operating system written in Erlang and Rust

SQL over RPC, safely

Native Rust runtime for adversarial extension workloads with deterministic replay, cryptographic decision receipts, and fleet-scale containment.

A scripting language for cowboy coders

A ground-up Rust microkernel operating system exploring intent-centric computing, capability security, semantic knowledge, and privacy-first system architecture.

A local, open-source capability gateway for AI agents — one AI-native self-describe endpoint so any agent can discover, understand, be granted, and call the software on your machine, under a trust model you can see, scope, and revoke.

Electron runtime layer providing protocol-based separation, component assembly, and capability-based process control.

Self-hosted AI agent for terminal and Telegram, with Claude/Codex workers, live activity and a deterministic permission kernel.

A statically-typed scripting language and bytecode VM for sandboxed execution of AI-generated code, built in C++ with no GC or JIT.

KAIROS-ARK is a high-performance, Rust-based Agent Runtime Kernel built for industrial-grade reliability. It delivers sub-100µs dispatch latency, event-sourced deterministic replay, and kernel-enforced capability sandboxing, bridging Python prototypes and production AI systems.

A programming language that reads like English — and can't touch your machine unless you say so. Capability sandbox, pattern matching, gradual types. pip install she-lang

An orchestration DSL and embedded runtime for hosting agentic systems. Durable, inspectable workflows with deterministic step ordering, running in-process under a capability jail — the model lives in your host, not in the runtime. 30+ companion packages for memory, transport and governance.

Execution governance for AI agents: signed action-bound decisions, Hosted effect mediation, replay resistance, and evidence-scoped isolation.

Agent-first display server: a small trusted core speaking a capability-native protocol, with every legacy app confined to its own per-app shim. Humans and AI agents operate the same GUIs under revocable, capability-scoped authorization.

A local-first runtime that runs AI agents behind a capability arbiter. Every tool call is gated against a declared manifest, scored on earned trust, and written to an immutable audit ledger.

A microkernel operating system written entirely in ManiT, the balanced ternary systems language. Trit-addressed memory with trit-trie page tables, three privilege rings one trit deep, photon-schedule capability security. Builds and runs on Linux today.