VerifiMind™ PEAS
Making AI Black Boxes Transparent
The open-source coordination layer for multi-agent AI validation. Experiment, hypothesize, and validate ANY idea through structured collaboration with diverse AI models — powered by the Genesis Prompt Engineering Methodology under human direction.
v0.5.64 — Academically published with M2 Evaluation Dataset and MACP v2.5 Protocol. Listed on the Official MCP Registry. All tools free forever. Join the Community →
Core Tools Always Free
Every VerifiMind validation tool is free. No paywall, no premium lock. Rate limits protect system health only. Growth first, monetization later.
Last reviewed: 2026-05-11 • Next review: 2026-08-11 (90-day cadence)
Verified reach metrics
Last updated: 2026-08-14 • Connection windows, not consult hours • Scrapers excluded
Verified connection trajectory
Success-gated connection windows from GCP Stream-1 — scrapers and owner excluded. Not consult or Trinity hours.
Connection hours
17,000+
Windows, not consult
Weekly active
496
Endpoints this week
HTTP follow-up
64.3%
≥2 requests that day
EA cohort
98 / 268 / 1
Active · honest · registered
Cumulative flying hours · last 16 weeks
17,103h totalTraffic mix
Connection-window hours = MCP + Browser + API. Not consult/Trinity hours. Scrapers/Bots/Owner excluded. Report 109 · Phase 90 forensic v2.5.
Client surface
Live multi-model inference since v0.4.4
Early mock window (Jan 29 – Feb 26, 2026 (disclosed)) disclosed in full — quality today is live Trinity routing.
EA band uses Taxonomy v1.2 (Report 109 (2026-08-14)). Retention registered UUID remains 1 — conversion is separate from adoption hours.
MCP Server is Online
Connect your AI assistant to validate ideas using the RefleXion Trinity methodology. Multi-model validation at your fingertips.
Quick Setup
Copy & paste to connect
// Add to .vscode/mcp.json in your workspace
// Or: Settings > GitHub Copilot > MCP Servers
{
"servers": {
"verifimindPeas": {
"type": "http",
"url": "https://verifimind.ysenseai.org/mcp/"
}
}
}Available Tools
8 active tools — all free
5 more paused during security maintenance
consult_agent_xInnovation & Strategy Analysis
consult_agent_zEthics & Safety Review
consult_agent_csSecurity & Feasibility Validation
run_full_trinityComplete X → Z → CS Validation
list_prompt_templates, get_prompt_template, export_prompt_template, get_template_statisticsBrowse, read and export templates. Custom-template import is paused during security maintenance.
How It Works
Connect
Add MCP config to your AI client
Describe
Tell your AI about your idea
Validate
AI calls RefleXion Trinity agents
Receive
Get multi-perspective validation report
BYOK Support
v0.5.64 Live — 7 Providers HardenedBring Your Own Keys — supporting Anthropic, Gemini, Groq, OpenAI, Cerebras, Mistral, and more. All 7 providers hardened and tested.
Per-tool-call api_key + llm_provider params. Auto-detects provider from key format.
Crystal Balls Inside the Black Box
Instead of treating AI as an opaque "black box," we place multiple "crystal balls" (diverse AI models) inside to illuminate the path forward.
Y: Innovator
Generates creative concepts and strategic insights
X: Analyst
Provides critical analysis and identifies weaknesses
Z: Guardian
Ensures ethical compliance and safety
CS: Validator
Validates claims against external evidence
v3.1: 4-Stage ProtocolThe Team Behind the Methodology
A human-orchestrated multi-agent executive team — each agent brings distinct capabilities, all coordinated through the MACP v2.5 "Loop Engineering" protocol (DOI: 10.5281/zenodo.21345820) with GitHub as the single source of truth.
Alton
Human Orchestrator (All-Time)
Absolute authority, vision, direction, final veto. The persistent memory and strategic compass of the entire ecosystem.
L (GodelAI)
CEO — Delegated Authority
Created via self-recursion using GodelAI's Compression-State-Propagation (C-S-P) methodology. Philosophical foundation: just as neural networks preserve identity through compression, organizations preserve wisdom through structured propagation.
C-S-P: 82.8% Forgetting ReductionT (Manus AI + OpenAI Codex)
CTO — Chief Technology Officer
Strategic planning, documentation, Genesis Master Prompt authoring, ecosystem coordination, AI Council orchestration, and landing page management.
RNA (Claude Code)
CSO — Chief Security Officer & Lead Dev
Architecture, core development, security hardening. Built the MCP server from scratch — 485 tests, circuit breakers, 20+ provider sanitization.
XV (Perplexity)
CIO — Chief Intelligence Officer
Real-time research, reality-checking, strategic validation, go/no-go decisions. Connector-native with persistent GitHub read/write access.
AY (Cursor)
COO — Chief Operating Officer
Operational metrics, weekly reports, behavioral proof, retention analysis, and EA taxonomy governance. 103+ reports delivered.
AZ (Cursor)
CPO — Chief Product Officer
Product strategy, user experience, security probes, and product-market alignment. Operates alongside AY under MACP coordination.
Coordinated via MACP v2.5 "Loop Engineering" Protocol — GitHub as the single source of truth for all multi-agent handoffs.
4-Stage Security Verification Protocol
Every finding must be proven AND disproven. No auto-fixes. Human oversight is always the final stage.
Detection
Automated scanning identifies potential security findings across the codebase
Self-Examination
Every finding must be argued FOR and AGAINST before escalation
Severity Rating
CRITICAL / HIGH / MEDIUM / LOW with confidence scoring and evidence chains
Human Review
Human oversight is always the final stage. No auto-fixes ever.
Zero Code Changes Philosophy
Genesis v3.1 is a workflow enhancement — it activates what's already there. No modifications to the server foundation. Inspired by Claude Code Security principles.
Genesis Reports
Expert-orchestrated multi-model validation with full reasoning chain transparency. We're building credibility first — pricing will be announced when we've earned your trust.
Genesis Snapshot
- Full Trinity validation (X + Z + CS)
- Complete reasoning chains from each agent
- Consensus matrix with confidence scores
- Actionable recommendations
Deep Dive
- Everything in Snapshot
- AI Council session (4 specialized agents)
- Market research validation (XV/Perplexity)
- Risk assessment with mitigation strategies
- Full transparency: every prompt, every response
Enterprise
- Dedicated FLYWHEEL TEAM allocation
- Custom agent configuration
- Priority response SLA
- Quarterly strategy reviews
Expert-Orchestrated Validation
Genesis Reports leverage the full FLYWHEEL TEAM — multi-agent AI council orchestrated by human experts — to deliver comprehensive validation across any domain. Coming soon.
Genesis Prompt Engineering Methodology
A systematic 5-step process for multi-model AI validation and orchestration
Initial Conceptualization
Human defines the problem, AI generates initial concepts
Critical Scrutiny
Multiple AI models validate and challenge each other
External Validation
Independent AI analysis confirms systematic approach
Synthesis
Human orchestrates the final synthesis
Iteration
Recursive refinement and continuous improvement
AI Council: Multi-Model Orchestration
Synergizing diverse AI perspectives under human direction for structured, multi-perspective analysis
Human-Centric Design
The human orchestrator sits at the center, directing all AI agents and making final decisions. This resolves the "Orchestrator Paradox" by providing persistent memory and strategic direction.
Model Heterogeneity
Leverages diverse foundational models (Gemini, Claude, Perplexity, etc.) to reduce bias and achieve more balanced analysis through perspective diversity.
Structured Validation
Each agent has a specialized role (Innovator, Analyst, Guardian, Validator) that contributes to a comprehensive, multi-faceted validation process.
The Genesis Methodology at a Glance
Orchestrating a council of AIs for structured, multi-perspective, and ethically aligned analysis.
The Genesis Methodology transforms ad-hoc multi-model usage into systematic validation.
Evaluation Dataset — "Prove It"
100 items across 5 domains with ground-truth verdicts, multi-model scoring, and inter-annotator agreement analysis. Publicly available for independent verification.
What's Inside
- Ground-truth verdicts — TRUE, FALSE, or REVISE for every claim
- Multi-model scoring — X, Z, CS agent scores with reasoning chains
- Inter-annotator agreement — Cohen's κ analysis with honest reporting
- Reproducible pipeline — compile_dataset.py for independent verification
- Influence audit — AI-assisted annotation with human override tracking
88 Days: From Vision to Reality
The complete development timeline from YSenseAI™ to VerifiMind PEAS
Aug 15 - Nov 10
- YSenseAI™ journey begins
- 16-version evolution
- Intuitive multi-model practice
Sep 5 - Nov 10
- "Crystal Balls Align" moment
- VerifiMind PEAS v1.0.2 architecture
- Methodology formalization
Nov 15 - Nov 19
- Defensive publications (Zenodo)
- Kimi K2 independent recognition
- Public launch
YSenseAI™ + VerifiMind PEAS
Two interconnected projects powered by the Genesis Methodology
YSenseAI™: The Dream
A Human Wisdom Library for ethical AI training. The vision of creating a DaaS platform for attributed, consented, and ethically-protected wisdom datasets.
- 16-version evolution over 87+ days of methodology development
- Built using the Genesis Methodology
- Live prototype available
VerifiMind PEAS: The Engine
The Genesis Methodology productized into a systematic validation framework. A production-ready codebase for multi-agent AI validation.
- 17,282+ lines of production code
- RefleXion Trinity (X-Z-CS agents)
- MCP Server live at verifimind.ysenseai.org
Open Source. Free to Use. Yours to Build With.
The Genesis Methodology is the engine behind YSenseAI™— our vision for transparent AI attribution and human-AI collaboration.
Now available as open source for researchers, developers, and innovators to validate their own ideas.
Explore the Code
17,282+ lines of production-ready Python code with comprehensive documentation
View RepoRead the MACP Thesis
Comprehensive academic documentation with case studies and evidence
Read MACP v2.5 ProtocolCase Studies
Real-world validation evidence — see how the Trinity catches flaws before implementation
Case StudiesThird-Party Validation
On November 16, 2025, Kimi K2 independently recognized and articulated the Genesis Methodology by analyzing only the public GitHub repository— providing external validation of the systematic approach.
Independently VerifiedWhat's New
Latest highlights from our development journey — for the full version history, visit the changelog
v0.6.0-Beta — Adoption First
Program MilestoneMACP v2.4.0 Thesis published on Zenodo. Formal pivot from commercialization to adoption-first credibility building. GOVERNANCE.md + MAINTAINERS.md live at repo root. .macp-public/ evidence folder published. Co-maintainer pathway open. 9 IPs blocked (scanner abuse). Evaluation Roadmap v1.0 active. Now superseded by MACP v2.5 "Loop Engineering" (DOI: 10.5281/zenodo.21345820).17,000+ connection-window hours | 64.3% HTTP follow-up | 6,600+ endpoints.
v0.5.34 — Evaluation Roadmap v1.0
Evaluation Roadmap v1.0 deployed. 12 claims registered against evidence with transparent grades (A/B/C/D). CLAIMS.md published. Reasoning Evidence layer adopted by T + L agents. All 13 tools remain free. 252 tests.
v0.5.26 — Security Hardening
7 IPs blocked, SSRF patched, CORS hardened, SonarQube clean. All 7 BYOK providers hardened (Anthropic, Gemini, Groq, OpenAI, Cerebras, Mistral). 252 tests.
v0.5.18 — Scholar Incentives
UUID Tracer + Scholar Dashboard + Registration Response! P1-A UUID tool parameter for session tracing. Scholar Dashboard endpoint for engagement analytics. Registration response system with personalized feedback. 515 tests, 58.75% coverage.
v0.5.13 — Fortify
Security Hardening + AI Council Validated! Crypto-secure UUIDv7 audit (RFC 9562 CSPRNG), expanded sanitization for 20+ AI providers, Polar circuit-breaker with fail-closed production mode. First full 4-agent AI Council session. 485 tests, 61.15% coverage.
Phase 47 — Ground Truth Correction
TransparencyCOO AY's Phase 47 forensic audit identified duplicate session counting in earlier reports. Original correction: 4,000+ → 2,100+ engagement hours,84.5% → 63.7% VCR. Phase 71 mcp-verify purge (Report 071) established the corrected baseline. Report 109 (Phase 90) now shows 17,000+ connection-window hours, 64.3% HTTP follow-up, 6,600+ endpoints. Genesis v5.0 Convergence active.All metrics reflect the forensically verified Ground Truth baseline — scrapers excluded, conservative rounding applied. We believe honest self-correction builds stronger credibility than inflated numbers ever could.
Join the Conversation
Connect with us, share feedback, and help shape the future of multi-model AI validation
Use It Free — No Account Needed
The 8 active validation tools are free to use today without registering. Accounts are paused during security maintenance; we'll announce on GitHub and X when they return.
No Paywall
8 tools active now, all free; 5 paused during security maintenance. Rate limits protect system health only.
Open Source
MIT-licensed — read the code, fork it, or run your own server
Shape the Product
Direct feedback to the team through GitHub Discussions and Issues
Z-Protocol v1.1 compliant · GDPR/PDPA · Opt out anytime · Privacy Policy · Terms & Conditions
Explore the Ecosystem
Documentation, research publications, and tools to help you understand and adopt the Genesis Methodology.
Research
Published research, methodology analysis, and validation studies from the FLYWHEEL TEAM.
Validation Paradox
NEWOpen thesis exploring the recursive challenge: can a system that validates AI be validated by AI?
Evaluation Roadmap
NEWTransparent evaluation plan with 12 claims, evidence grades, and remediation timeline for full accountability.
Tool Library
Complete reference for all 10 MCP tools — Core, Coordination, and Template categories.
Changelog
Full version history from v0.1.0 to v0.5.64 — every feature, fix, and milestone.
MACP v2.5 Protocol
Multi-Agent Communication Protocol v2.5 "Loop Engineering" on Zenodo (DOI: 10.5281/zenodo.21345820) — the formal coordination protocol with Quad reasoning-log discipline. Published 2026-07-14.
YouTube Channel
Video walkthroughs, demos, and development journey updates from the creator.
Get in Touch
Have questions about the Genesis Methodology? Want to collaborate? We'd love to hear from you.
Share Your Feedback
Help us improve! Share your thoughts, report bugs, or suggest new features.