tokyo
Framework for building agent-friendly CLIs
Open-source projects. Clever little tools. Big ideas.
Unearthed from the Hacker News community.
Framework for building agent-friendly CLIs
Claude-real-video - any LLM can watch a video
Cosmonapse – an event-driven protocol for AI agent systems
CorvinOS – self-hosted agentic OS where EU AI Act and GDPR compliance by design
Open-ultra: a self-training LLM routing proxy
Mincdp – Give Claude eyes and hands in a browser, in ~410 lines of C
FleetView – a browser cockpit for running many Claude Code agents at once
AegisDB – self-hosted memory for AI agents, in one C binary
Yes-Brainer: a councils of LLMs debating to a consencus (open-sourced, BYOK)
Shiploop – an autonomous delivery loop for Claude Code, pitch to MVP
Veta: AI agent that QA-tests Android apps
Cc-session-migrate: Cross-node migration for Claude Code sessions
Ravenspire – watch your Claude Code and Codex agents as a JRPG
QuantmLayer – kernel-enforced containment for AI coding agents
Muxer – open-source model multiplexer for Claude Code
I made ThoughtDAG – LLM as an editable graph, wires are the context
Piggy – lazy senior dev mode for AI agents (80–94% less code)
ButterClaw – AI agent runtime security, SIGKILL on breach, no cloud