Python and TypeScript SDKs for verifiable evidence of AI agent actions. Signed receipts, policy enforcement, audit trails. Works with LangChain, CrewAI, MCP.
AI 执行与监察体系(Agent Oversight Framework):分层治理规范 + L2 运行时监察参照实现——证据分级、自校正、边界诚实。A layered governance spec for AI agents, plus a runnable reference implementation of the L2 runtime-oversight layer. Docs CC BY-SA 4.0 / code Apache-2.0.
CUMCM 数学建模竞赛全流程实战技能手册:读题自查→建模出图→LaTeX论文→交叉审阅→填 result 模板→AI 自查表终审→清理打包。含题意理解红线、数值严谨性守则、去 AIGC 特征清单、2026 AI 合规与 26 条自查终审框架。