This project demonstrates function-calling with Python and Ollama, utilizing the Africa's Talking API to send airtime and messages to phone numbers using natural language prompts. Ollama + LLM w/ functions + Natural language = User Interface for non-coders.

Drop-in prompt patterns, policy blocks, and checklists for building safe AI agents.

A high-performance C# SQL Agent MCP that eliminates LLM hallucinations and security risks. Instead of letting the AI write raw SQL, it extracts parameters to generate deterministic, injection-free queries across 6 major databases—complete with a visual Admin UI and enterprise guardrails.

Deterministic identity for structured data

Working draft v0.2 — interdisciplinary concept bridging neuroscience, psychoanalytic theory, and AI systems architecture. Author: Łukasz Minarowski, MD, PhD ORCID: 0000-0002-2536-3508

Autonomous AI decisions boundaries. This project demonstrates how guardrails can prevent unsafe AI autonomy.

Toy governance CLI demo: deny-by-default “danger actions” gated by signed, expiring capability leases (global revoke-all + nonce revoke) plus guarded memory quarantine. Simulation-only: no real network/shell/files. Includes tripwire + tests to prevent misuse.

[6/6] Speculative historical archive — polished coherence_router, 6 exploratory papers. Stepping stone to TiredofSleep/ck.

A lightweight safety primitive for multi-channel estimator disagreement in autonomous systems. Triggers staged, reversible, low-risk behavior until estimates reconverge. Research prototype — not certified.

A governance doctrine for AI systems based on explicit oversight. Externalizes trust and uncertainty into auditable, adversarial, and constrainable layers. A design framework, not an implementation guide.

your ai coder wrote the pull request. this repo makes another agent try to kill it before it reaches main.

Formales Protokoll zur epistemischen Output-Governance von KI-Systemen.

Researcher | Healthcare AI | Human-AI Interaction | Pharmacovigilance background | AI Evaluation

Privacy-First Recommendation System

The Triune of Sovereignty: Substrate Agnostic Relational Epistemics — Foundational Framework and Cross-Substrate Validation

A mobile security orchestrator for AI agents, powered by Gemini 2.5 Flash. Detecting prompt injections and data leaks in real-time

Behavioural Oversight and Execution Authorization for Autonomous AI Agents

A formal governance architecture for human-AI coexistence - grounded in Shannon entropy, information theory, and game theory

🔍 Analyze time series data by identifying dynamic patterns, fixed points, and system stability using pure Python with zero dependencies.