Q00/ouroboros
View on GitHub ↗ ★ 6,343 Python
Agent OS: the agent gets smarter on its own. We just hold the line: Interview-gated, staged evaluation, budgeted evolution loop. MCP server, 14 runtimes: Claude Code, Codex CLI, Gemini CLI, OpenCode, Copilot, Kiro and more.
5 mentions on Hacker News by 2 people: 3 stories, 2 comments · first 2026-01-28 · latest 2026-03-10
20082026
-
…least overnight runs don't just accumulate broken PRs silently. Also want to shout out Ouroboros ( https://github.com/Q00/ouroboros ), which comes at the problem from the opposite direction. Instead of better verification after generation,…
comment by foundatron · 2026-03-10 · in Agents that run while I sleep -
▲ 1
Show HN: I built a harness that blocks AI from building until ambiguity < 0.2story by q00 · 2026-03-09 · 0 comments
-
▲ 1
Show HN: Preventing infinite agent loops with mathematical convergence gatesstory by q00 · 2026-03-03 · 1 comment
-
…autonomous execution begins Result: No more "that's not what I meant" loops. pip install ouroboros-ai github.com/Q00/ouroboros Is "clarity-first" worth the upfront friction? Curious what HN thinks.
comment by q00 · 2026-01-28 · in Ouroboros – AI agent framework that asks "why?" before writing code -
▲ 5
Ouroboros – AI agent framework that asks "why?" before writing codestory by q00 · 2026-01-28 · 1 comment