Agents

Experimental Python framework for a self-building autonomous agent, centered on stored functions and a dashboard

Local desktop app that wraps Codex and Claude Code with a shared browser, a file viewer and one conversation across agents.

Open-source agentic language model (30.5B parameters, 3.3B active per token) trained for long, multi-step web information seeking.

Personal agent with a Bun gateway that keeps memory, skills and run history, with approval-gated tools and web, CLI and mobile clients.

Browser workbench that coordinates Claude Code, Codex, Gemini, OpenCode and other CLI agents as a visible team of local processes.

Open-source DevOps agent that runs 24/7 on your own machines, works on schedules and messages you only when it needs a person.

Browser automation agent that runs in a Chrome side panel and carries out web tasks with OpenAI or Anthropic models.

Open-source, local-first workspace where humans and AI coding agents coordinate through channels, tasks and persistent memory.

Sakana AI framework for automated research where LLMs generate ideas, run experiments and write papers.

Self-hosted AI agent built on n8n, PostgreSQL and Claude that runs over Telegram or a webhook API.

Python command-line pairing agent for ML engineers that builds baselines, works Kaggle-style tasks and writes weekly reports.

Hub that launches Claude Code, Codex CLI, Gemini CLI and other agents with third-party APIs and reaches them from web, mobile and chat apps.

Modular LLM software engineering agent from ByteDance with a CLI, interactive mode and several model providers.

Multi-agent research workbench that searches arXiv, OpenAlex and Semantic Scholar and drafts evidence-grounded literature reviews.

Python port of the opencode coding agent with a CLI, a FastAPI server and MCP client support.

Local desktop and web workbench that reviews a repo's pull requests with Claude or Codex in isolated worktrees and posts only approved comments.

Go reimplementation of the pi coding agent with headless and REPL modes, skills, plugins and multi-provider support.

Claude Code agent on a VPS you own, with a real Chromium browser and a Telegram bot to message it.

Local-first AI coding terminal that grounds model output in real files, tools, permission checks and verification steps.

Local-first desktop workbench that runs several coding agents in separate git worktrees and handles PRs, merges and CI fixes.

Desktop AI agent that runs on your machine, edits project files and runs commands, and produces Word, Excel, PowerPoint and PDF files.

Open-source terminal coding agent written in Rust, an independent implementation inspired by Claude Code

Self-hosted AI agent harness for authorized penetration tests and security labs, using your own model API keys.

Open-source deep research agent for multi-step web research, with models published on Hugging Face.