Agents

Strip AI provenance marks from text and files you own, with an unusually honest accounting of what actually disappears and what merely gets rewritten.

A self-hosted gateway that taps WeChat through an undocumented protocol to run AI roleplay bots with personas, stickers, and visual chat flows.

An interactive tutorial that rebuilds the 'pi' coding agent in ~600 lines of TypeScript, revealing the source code line-by-line as you read.

Phone Harness turns macOS iPhone Mirroring into a stateless bridge so an LLM can see, tap, and type on a real iPhone without jailbreaks, Xcode, or WebDriverAgent.

Agent-side workflows that turn novels into production-ready short-drama bibles—characters, outlines, and art guides—without leaving Claude Code or Codex.

Cloudflare Computer gives AI agents a persistent, SQLite-backed workspace that can run shell commands, JavaScript, or full Linux containers without syncing state between them.

It stops long-running agents from drifting by forcing them to verify progress through independent roles before every new round.

Brigade is a self-hosted, terminal-first runtime that coordinates persistent AI agents through an org chart and shared long-term memory.

Flint compiles compact, human-editable chart specs into polished Vega-Lite, ECharts, or Chart.js visualizations so agents and developers can skip hand-tuning scales, axes, and spacing.

A Python layer that makes AI agents explain themselves through structured context graphs, decision trails, and W3C-compliant provenance.

LoopX is a local state kernel that keeps long-running agent objectives, evidence, and handoffs stable across sessions, tools, and reboots.

Macro exists because context dies when your email, chat, docs, and tasks live in separate apps, so it links them into one queryable workspace with shared AI memory.

This project exists because AI-generated Chinese is fluently anonymous, so it encodes hard editorial rules and a prose linter to force models to write with the specific gravity of a real person.

It gives AI dev agents a persistent memory and long-term runtime on your hardware so they learn, schedule tasks, and resume work instead of starting fresh every chat.

Ante is a self-contained Rust agent core designed to run thousands of lightweight, model-agnostic coding agents without runtime dependencies or vendor lock-in.

Waku unifies Claude Code, Codex, and several other agent CLIs into a single native desktop app that keeps all transcripts and state local.

iFixAi runs up to 32 inspections against any LLM or agent and returns a letter-grade scorecard in minutes, using a separate provider as judge so the model isn't grading its own homework.

MAC turns a text prompt into a printable 3D model by dividing the work among four specialist LLM agents, cutting token consumption by 116× compared with monolithic approaches.

Why wire agents by hand when an LLM can build, benchmark, and evolve them for two cents?

A hybrid CLI tool that uses deterministic pipelines to keep LLM agents from drifting off-target during code review.