A distributed research lab advancing open-source AI agent architectures and high-performance systems engineering.
Autonomous coding agents, tool-use synthesis, and agent mesh architectures for next-generation AI systems.
Sub-millisecond frameworks, API gateways, and distributed infrastructure primitives designed for extreme throughput.
MIT-licensed components you can use, modify, and contribute to freely. Research published in the open.
Open-source projects spanning AI agents and systems infrastructure — explore the full collection on GitHub.
Depth-per-dependency as a defensive architecture. A five-module supply chain, a fails-closed command classifier, and a default-on sandbox — compared against Claude Code and Codex using each vendor's official security documentation.
Why security-conscious infrastructure is the defining requirement of the agent era. A Trojan hides in 100 lines of plausible code, the first AI-driven end-to-end breach just happened — and defense-in-depth below the model is the only answer. How odek aims to be tomorrow's most secure general-purpose agent.
Why the future of software operations is not more monitoring, but better observability, verification, and autonomous remediation. 42% of committed code is already AI-generated, but 96% of developers do not fully trust it. AIOps is becoming the load-bearing infrastructure of AI-native engineering.
The Lights-Out Software Foundries Reshaping the World by 2030. How autonomous agent meshes, closed-loop verification, and lights-out infrastructure are turning software production into a utility — and why the real question is whether we can still govern the factories we build.
An orchestration harness for OpenCode. A fixed request lifecycle routes work to specialist agents, runs review, testing, and documentation in parallel, enforces quality gates, and keeps persistent per-project memory across sessions. A configuration layer over OpenCode — repeatable, inspectable software development.
High-performance Go static file server built on fasthttp. Delivers ~148k req/sec — 59% faster than Bun's native static server. Features near-zero alloc hot path, gzip/Brotli/Zstd compression, TLS 1.3, smart LRU caching with startup preloading, HTTP/2, byte-range serving, CORS, and security hardening (path traversal prevention, HSTS, CSP). Powers static.21no.de.
Modern TypeScript dependency injection container optimized for Bun runtime. Blazing fast with sub-microsecond resolution (~26M ops/sec for cached singletons), lightweight (<10KB), and fully type-safe. Supports singleton, transient, and scoped lifetimes with flexible API options: decorators, functional, or imperative. Built-in security against prototype pollution, memory leaks, and DoS attacks. Includes testing utilities.
Next-generation HTTP gateway and load balancer harnessing the speed of Bun. Enterprise-grade features with zero-config simplicity.
Lightweight and fast Node.js framework for building RESTful APIs with Express-like syntax and superior performance. Includes built-in support for middleware, routing, request/response handling, and error management. Features comprehensive documentation, extensive examples, and active community support. Designed for microservices, serverless functions, and high-throughput applications.
Super-fast Node.js API gateway with built-in rate limiting, circuit breaker, service discovery, and load balancing. Supports multiple protocols and authentication strategies. Ideal for microservices architectures requiring minimal latency and maximum reliability.
Tiny, blazing-fast Node.js HTTP router with zero dependencies. Sub-microsecond routing performance, lightweight footprint, and a clean API. Ideal for microservices, serverless functions, and embedded HTTP servers where every microsecond matters.
Zero-dependency Model Context Protocol server framework for Go. Build AI-accessible tools, resources, and prompts over stdio JSON-RPC. Single binary, no runtime, sub-millisecond cold start. 8 MCP methods, 97.8% test coverage, Docker E2E pipeline. The first Go-native MCP framework.
Zero-dependency vector similarity library for Go. Cosine, Euclidean, Manhattan, and Dot Product distance metrics with brute-force nearest-neighbor search. Built-in text embedding via sparse random projections (Johnson-Lindenstrauss). Gob and JSON persistence with full roundtrip. 40 tests, 96.8% coverage.
Thoughts on AI, systems engineering, and the future of infrastructure. Updated regularly on Medium.