AI Agents
moltis
Secure persistent personal agent server in one Rust binary — sandboxed, multi-provider, on your hardware
★ 2.9k starsMIT
View guide →An Agent OS with interview-gated, budgeted evolution — one MCP server across 14 agent runtimes.
ouroboros is an open-source ai agents project by Q00 (Q00/ouroboros on GitHub). It has 6,195 GitHub stars and 623 forks, is released under the MIT license, and was last updated on GitHub on 7 October 2026.
Prepare the idea you will hand to the interview — Ouroboros starts by questioning you before any building happens.
Use the one-line installer for your platform, or install the package with pipx/uv.
Run the setup for your agent runtime (Claude Code, Codex, Kiro, Copilot and more) so the MCP server is registered.
Start with ooo setup and ooo interview, answer the gating questions, and let the staged, budgeted loop build and evaluate the result.
Nothing is built until an interview pins down intent — the system questions you first, then executes against that contract.
Agents evolve solutions over up to 30 generations with convergence checks and stagnation detection, under an explicit budget.
Work passes staged evaluation with a weighted drift measure across goal, constraints and ontology before it is accepted.
The same workflow runs on Claude Code, Codex CLI, Gemini CLI, OpenCode, Copilot, Kiro, Hermes, Goose and more via pluggable runtime backends.
Routes work across Frugal, Standard and Frontier model tiers with auto-escalation on failure and auto-downgrade on success.
Capability discovery, policy, directives and an event journal make agent execution replayable and auditable.
Clone Ouroboros if you want the source; most users use the installer in the next step.
git clone https://github.com/Q00/ouroboros.git
cd ouroborosRun the official installer for macOS/Linux/WSL2 (a PowerShell installer exists for Windows).
curl -fsSL https://raw.githubusercontent.com/Q00/ouroboros/main/scripts/install.sh | OUROBOROS_INSTALL_REF=readme-hero bashInitialize Ouroboros and start your first interview-gated build.
ooo setup
ooo interview "I want to build a task management CLI"Ouroboros is maintained by Q00 and contributors in the Q00/ouroboros repository.
Yes — it is MIT licensed. You pay for the model usage of the runtime and models you connect.
It drives external agent runtimes and models, so you need working access to at least one supported runtime (for example Claude Code or Codex); Ouroboros itself charges nothing.
A plain agent starts coding immediately; Ouroboros gates work behind an interview, evaluates in stages, and evolves solutions within a budget — trading speed for control and repeatability.
Run the one-line installer, then ooo setup for your runtime and ooo interview with your idea.
Secure persistent personal agent server in one Rust binary — sandboxed, multi-provider, on your hardware
Exa's MCP server for web search and crawling — give your AI agent a live, high-quality web index.
Monid — OpenRouter for agent tools: one base URL and one key give an agent 2,000+ tools across 72+ providers
Open Cowork: a desktop AI agent app with one-click Claude Code, MCP tools, skills, and VM sandboxing.
Mobile app automation and verification built for AI coding agents — CLI, MCP server, and Node API
A token-conscious MCP server that connects AI tools to Postgres, MySQL, SQLite and more.
The most-starred other AI Agents projects in the LibreAI catalogue — the realistic shortlist if ouroboros is not the right fit.
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
Free and open — grab the source, star the repo, and start creating.