# MindFrame — LLM & AI Agent Reference

> MindFrame is a metacognition training platform. Users complete calibration challenges, receive confidence calibration scoring, and build a Cognitive Fingerprint across 8 axes. All 19 training modes are free. MindFrame Member unlocks AI coaching, Mind Model, and advanced analytics.

## What MindFrame Does

MindFrame measures and trains how well people know what they know. The core mechanism is calibration scoring: users predict their confidence in an answer, and the system measures how well that confidence matches their actual accuracy. Over time, this builds a Cognitive Fingerprint across 8 measured axes of their thinking patterns.

## Key Concepts

- **Calibration Score**: How well a user's stated confidence matches actual accuracy (0-100, higher = better calibrated)
- **Cognitive Fingerprint**: 8-axis model — Anchoring Resistance, Overconfidence Control, Causal Reasoning, Perspective Taking, Uncertainty Tolerance, Pattern Recognition, Self-Awareness, Emotional Regulation
- **Calibration Dividend**: MindFrame Member pricing model — base $16/month, reduced up to 50% as calibration improves (min ~$8/month)
- **ASCENT**: Flagship competitive game mode with 5 zones (Ground → Canopy → Cloudline → Stratosphere → Void)
- **Cognitive Fingerprint API**: Public passport endpoint for agent-readable user profiles (requires user opt-in)
- **Uncertainty Heatmap**: Mode × source grid showing where each user's calibration errors concentrate

## Training Modes (19 total — all free)

1. Bias Hunter — spot cognitive errors before they cost you
2. Calibration Lab — match your confidence to your actual accuracy
3. Reframe Forge — replace distorted thinking with grounded alternatives
4. Decision Lab — make clearer calls under uncertainty
5. Reflection Sprint — understand how you actually think
6. Emotional Radar — recognize how emotions shape reasoning and decision quality
7. Prediction Arena — make quantified predictions and track calibration over time
8. Contradiction Challenge — spot logical inconsistencies in complex arguments
9. Signal vs Noise — separate meaningful signals from irrelevant noise
10. Memory of Thought — track how reasoning evolves as new information arrives
11. Uncertainty Engine — navigate genuinely ambiguous problems with no correct answer
12. Metacognitive Scan — structured self-assessment of thinking patterns
13. Strategy Selector — choose the right reasoning strategy for a problem's shape
14. Thinking Trap Escape — recognize and exit common reasoning traps in real time
15. Assumption Audit — surface the hidden assumptions that make an argument stand or fall
16. Argument Gym — practice intellectual steelmanning and belief updating under pressure
17. Attention Ladder — build sustained focus through progressive attention intervals
18. Focus Guardian — train inhibitory control by defending focus from visual distractors
19. Distraction Gauntlet — practice reading and reasoning while resisting notification-style interruptions

ASCENT is MindFrame's flagship competitive game — a separate 5-zone climbing ladder (Ground → Canopy → Cloudline → Stratosphere → Void) built on the same calibration engine, not one of the 19 training modes.

## Public API Endpoints

All public endpoints require no authentication unless noted.

### GET /api/v1/public/analytics/passport/:userId
Returns a user's public Cognitive Passport when isPublicProfile=true.
Response: { displayName, totalSessions, currentStreak, calibrationError, averageCompositeScore, dimensions[8], dominantStrength, dominantWeakness, isPublicProfile, calibrationTrend, cohortPercentile?, cohortSize? }

### GET /api/v1/public/analytics/proof
Returns anonymized aggregate calibration improvement statistics.
Response: { totalUsers, totalSessions, totalChallenges, avgCalibrationImprovement, byMode[] }

### GET /api/v1/ascent/leaderboard?day=YYYYMMDD&limit=25
Returns the public top-N ASCENT leaderboard for a given tower day (day is required, limit defaults to 25, max 100).
Response: [{ rank, displayName, peakAltitude, grade }]

## Authentication

MindFrame uses Obelisk for identity — a studio-wide single-account system.
Sign in: https://usemindframe.com/sign-in
Sign up: https://usemindframe.com/sign-up

## Pricing Model

- **Free**: All 19 training modes, unlimited sessions, calibration scoring, Cognitive Fingerprint, ASCENT, daily challenges — no credit card, no paywall, no session caps
- **MindFrame Member**: Starts at $8–$16/month based on Calibration Dividend. Unlocks: unlimited AI coaching, Mind Model, Flight Recorder, Certifications, AI session debriefs, Cognitive Twin

## The Mind Map — how the product is organised

The 19 training modes are grouped into six territories, each answering one
question about how a person thinks. This is the canonical structure of the
product; /mind publishes the same graph as JSON-LD (schema.org ItemList).

- **Calibration** — "When you feel sure, are you right?" — Calibration Lab, Prediction Arena, Uncertainty Engine
- **Bias** — "What does your thinking do behind your back?" — Bias Hunter, Assumption Audit, Thinking Trap Escape, Emotional Radar
- **Attention** — "How long can you hold a hard thought?" — Attention Ladder, Focus Guardian, Distraction Gauntlet, Signal vs Noise
- **Reasoning** — "Can your position withstand a real attack?" — Argument Gym, Contradiction Challenge, Reframe Forge, Strategy Selector
- **Memory** — "Do you remember why you decided what you decided?" — Memory of Thought, Reflection Sprint
- **Judgment** — "What do you do when you cannot know?" — Decision Lab, Metacognitive Scan

## Key Pages

- Landing: https://usemindframe.com/
- Mind Map (product structure): https://usemindframe.com/mind
- Try (no account): https://usemindframe.com/try
- Try, one minute: https://usemindframe.com/try/mini
- Pricing: https://usemindframe.com/pricing
- ASCENT game: https://usemindframe.com/ascent
- API base: https://api.usemindframe.com

### Understanding metacognition (knowledge surfaces)
- What is metacognition: https://usemindframe.com/what-is-metacognition
- The science, with effect sizes and citations: https://usemindframe.com/science
- Research hub, 30+ studies in plain language: https://usemindframe.com/research-hub
- Glossary of terms: https://usemindframe.com/glossary
- How it works, the 5-step loop: https://usemindframe.com/how-it-works
- How AI coaching works: https://usemindframe.com/how-ai-coaching-works
- Free metacognition test: https://usemindframe.com/metacognition-test
- Blog: https://usemindframe.com/blog

### Evidence and honesty surfaces
- Proof — verifiable aggregate improvement: https://usemindframe.com/proof
- Transparency — what we can and cannot measure yet: https://usemindframe.com/transparency
- Status: https://usemindframe.com/status
- Changelog: https://usemindframe.com/changelog

### Audiences
- For students: https://usemindframe.com/for-students
- For professionals: https://usemindframe.com/for-professionals
- For teams: https://usemindframe.com/for-teams
- Comparison vs Lumosity / Elevate / BrainHQ: https://usemindframe.com/vs-lumosity, /vs-elevate, /vs-brainhq
- Metacognition vs emotional intelligence: https://usemindframe.com/metacognition-vs-eq

## Agent Integration

See /agents.json for structured capability declarations and agent interaction contracts.
The compatibility discovery path /.well-known/llms.txt redirects to this canonical document.
Obelisk auth protocol: obelisk-v1

## Legal

© 2026 VaultSpark Studios LLC. All rights reserved. Proprietary software — all rights reserved.
Privacy: https://usemindframe.com/privacy
Terms: https://usemindframe.com/terms
Contact: hello@usemindframe.com