decision-framework

二战四巨头决策智慧技能包 — 罗斯福·实验性治理 / 丘吉尔·动员叙事 / 斯大林·权力工程 / 蒋介石·韧性重建

StormGrill is an open-source dual-protocol orchestrator that pairs verified multi-agent research (Storm) with real-time adversarial interrogation (Grill) to help decision-makers stress-test strategic decisions against facts, not intuition. Make bad strategic decisions expensive to make.

Socius — Your AI Work Partner. 认知 + 决策 + 执行三大引擎的统一 Runtime。不是工具链，是能替你工作、和你并肩作战的 AI 搭档。

Research codebase for simulation-oriented analysis of multi-model LLM evaluation workflows in Library and Information Science.

A self-contained, AI-orchestrated decision-making and problem-solving framework. Point any AI assistant at the folder + tell it to read AI-BOOTSTRAP.md.

Public-data-calibrated decision pipeline for privacy-preserving agentic replenishment in retail supply chains.

Normative CHP v1.0 spec + conformance suite — adversarial decision hardening, R0 gate, domain score floors, signed lock states

Applied-AI build-vs-buy study: benchmarking 6 GenAI models (OpenAI/Anthropic/Google) vs a fine-tuned RoBERTa specialist on CFPB banking-complaint root-cause classification. UNC Kenan-Flagler MBA742 research project.

AI agent methodology navigation skill — 35-methodology index map with routing tables, invocation protocols, degradation paths, and correction mechanisms for choosing the right framework first

A formal multi-agent LLM framework for decision-making, AI evaluation, and uncertainty analysis using structured divergence.

Velocity decision tool for specialty food brands. 8 decisions — shelf defense, production planning, promo ROI, distribution expansion, SKU rationalization, launch diagnostics, pricing power — built on a synthetic 50-SKU, 1.2M-row retail dataset.

Within- and across-candidate non-predictive diagnostic analysis of rubric-based profiles using interpretable statistical structure.

Decision and execution framework that sits above any LLM to improve judgment, clarity, and sales decisions.

Developer and enterprise infrastructure for building hardened, human-auditable multi-agent decision workflows.

MW-Ω v2.2 — Meta Workflow operating system for deterministic institutional governance