Skip to content
#

thinking-protocol

Star

Here are 5 public repositories matching this topic...

Language: All
Filter by language
All 5 HTML 1 JavaScript 1

263311487-ux / falsify

Star 0

The scientific thinking protocol for AI agents — falsify before you believe. 5-stage falsification skill (axioms → hypothesis → adversarial → verify → converge) for Codex, Claude Code, Cursor, Gemini CLI & 20+ agents. Dual-model evals 26/28. 五段式科学思维协议：公理 → 假设 → 对抗 → 验证 → 收束。

skill calibration bayesian critical-thinking codex epistemology reasoning mental-models scientific-thinking ai-agent llm prompt-engineering falsification claude-code thinking-protocol
  • Updated Aug 30, 2026
  • HTML

Add this topic to your repo

To associate your repository with the thinking-protocol topic, visit your repo's landing page and select "manage topics."

Learn more