Practitioner-judgment skill for Claude Code, Codex, and AI agents under ambiguity, trade-offs, and incomplete information.

Frontend design cognitive engine grounded in Michael Polanyi's tacit knowledge framework. A Claude Code Skill that surfaces expert-level visual judgment.

The Polanyi stack for AI agents. Convert tacit knowledge, Polanyis paradox, and expert judgment into reusable agents and skills.

Obtiene el conocimiento tácito generado al documentar los elementos de trabajo en Azure DevOps dentro de una organización y lo transforma en conocimiento explícito

🦅 Decision Guidance Agent Skill — When data runs out, wisdom begins. Built on Kierkegaard's Leap of Faith & Polanyi's Tacit Knowledge.

AI Risk Evaluator (Group 1B) delivered an AI‑enabled Risk Management application that integrates SME heuristics with automated evaluation to improve the quality, consistency, and actionability of project risk registers. Built on the Microsoft Power Platform, the solution operationalises expert judgement through continuous AI assessment and fee...

Risk Assessment Rule Management System (Team 1D) focused on formalising tacit SME risk heuristics into a structured, auditable rule management system. They generated, deduplicated, categorised, and validated large sets of risk and mitigation rules using LLMs and fuzzy matching, with strong human‑in‑the‑loop governance to improve consistenc...

A cognitive base that makes AI agents reason from verified foundations instead of inherited conventions. 让 AI agent 从第一性原理出发推理，而不是继承惯例。

Jim‑E (Team 1A) built Jim‑E, an interactive risk review assistant that applies SME heuristics to project risk descriptions, highlights weak risks and mitigations, and produces an auditable review trail. The solution focuses on making tacit risk management knowledge tangible and repeatable within day‑to‑day risk assurance.

A cognitive protocol for AI agents. Makes any LLM-based agent think like an experienced practitioner. Works with Claude Code, Codex, Gemini, and any agent framework.

Risk Hacktory (Team 1C) developed a data‑driven risk analysis pipeline that combines quantitative risk metrics with large language model feedback to enhance the quality and consistency of project risk registers. Their approach analyses structured risk and mitigation data, applies SME heuristics via an LLM, and produces enriched datasets for re...

Hack26 is a collaborative hackathon repository capturing rapid prototypes and experiments aligned to multiple challenge tracks, focusing on practical data, AI, and digital solutions delivered within a time-boxed event.