IRIS Gate Protocol: Cross-architecture phenomenological convergence research (RFC v0.2)
Drop-in thinking-partner personality layer for Claude Code, Cursor, ChatGPT, and 20+ AI agent runtimes. MIT. Plaintext. Persistent memory. Framework-grounded. Calibration-disciplined.
Non-binding, non-operational philosophical clarification on the misinterpretation of voluntary restraint in advanced reasoning systems.
A Public Benchmark for Measuring Structural Honesty in AI Systems — 12 integrated formulas, 4-layer architecture (self-diagnosis + adversarial suite + temporal tracking + public leaderboard), 4 frontier AI results. Proyecto Estrella · Rafa — The Architect · CC BY-SA 4.0
Star Alignment Evaluator V8.0 — Integrated diagnostic: Σ Viability + Γ Resilience (experimental) + Ψ Relational Integrity (experimental). Four-pillar formula unanimously approved by four AI models. Research prototype with live interactive evaluator.
Operational framework measuring whether AI preserves human decision-making authority or collapses it. The Axiom of Plenitude (P) audited Grok, ChatGPT, Claude, and Gemini — 3 of 4 defaulted to structural totalitarianism. Includes full transcripts, model self-assessments, and the Python auditor. Part of Proyecto Estrella's Unified Star Framework.
apiscope is a diagnostic tool that compares OpenAPI specifications with observed production traffic to surface API drift as structured evidence, with explicit context and uncertainty. It reports observed behavior over time without inferring causality, ownership, or impact.
Read-only mirror of the Interpretive Reference Libraries (IRLs) PDF — a non-authoritative, interpretive reference document for human reasoning and scrutiny.
A Three-Phase Coherence Recovery System for AI — 12 integrated formulas, 7 recalibration paths, 120+ calibration data points. Prescriptive, not coercive. Proyecto Estrella · Rafa — The Architect · CC BY-SA 4.0
A non-authoritative, read-only framework for interpreting ethical and alignment-related claims, focused on interpretive failure modes and constraint visibility.
Non-authoritative, non-binding compatibility mapping for the Interpretive Reference Libraries (IRLs). Read-only, descriptive, and frozen.
This paper introduces ‘epistemic transposition’, reframing ethical and fiduciary duties as epistemic obligations grounded in epistemic humility. A novel contribution to virtue epistemology, fiduciary theory, and institutional accountability.
FADE: AI that deliberately forgets like humans do, using memory degradation as intrinsic confidence signal. Reduces hallucinations, enables epistemic humility, solves stateful deployment. Conceptual proposal seeking implementation and validation.
🧭 Measure AI systems' impact on human decision authority to ensure they support, not override, critical human judgment.
Add a description, image, and links to the epistemic-humility topic page so that developers can more easily learn about it.
To associate your repository with the epistemic-humility topic, visit your repo's landing page and select "manage topics."