# The Continuity Lab™ — MyShape Protocol · Full AI Reference > Version: 2026-07-19 | RFCs: 2 | Experiments: 576 | Tests: 121 > Open protocol research. All data public. MIT license. ## 1. WHAT THIS IS The Continuity Lab researches whether continuity — the unbroken chain of sovereign existence — can be made a verifiable cryptographic property. The MyShape Protocol is our first publication: a motion-signature system that verifies two sensor observations describe the same physically embodied entity. Core thesis: Identity verification answers "Who are you?" Continuity verification answers "Are you still you, across time?" The internet has the first. It has never had the second. ## 2. RFCs (FORMAL SPECIFICATIONS) ### RFC-0001: Motion Signature Format https://www.myshape.com/research/notes/004-motion-signature-rfc Defines the complete motion verification pipeline: - Presence Entropy Score (PES): 4D feature extraction (micro-timing variance, noise residual, frequency entropy, biological perturbation). PES = min(1, 0.25·MTV + 0.25·NR + 0.30·FE + 0.20·BP) - Jerk Peak Detection: MAD-based dynamic threshold (τ = median(|j|) + 2×MAD(|j|), floor 0.15 m/s³). 150ms refractory period. - Cross-Modal Temporal Matching: IMU jerk peaks matched to camera direction changes within ±500ms window, 90° direction tolerance. - Challenge-Response Extension (EE-003): 3 rounds, randomized directions, jittered timing (±1000ms). Evidence sufficient when ≥2/3 rounds pass. ### RFC-0002: Continuity Proof Format https://www.myshape.com/research/notes/006-continuity-proof-rfc Defines the proof infrastructure: - EvidenceReceipt: JSON-serializable object with sessionId (UUIDv4), timestamp, verdict, confidence, engineEvidence[], evidenceDigest (SHA-256), previousReceiptHash, cfcResults[] - Hash Chaining: R₁ → H(R₁) → R₂ → H(R₂) → R₃. Chain Integrity Score (CIS) = valid links / total links. - CFC Catalog: 8 Continuity Failure Conditions with severity levels and detection algorithms. - Verification Policy: acceptThreshold (0.70), rejectThreshold (0.35), IMU_ONLY_CAP (0.65), escalationEngines ([EE-003]) ## 3. CONTINUITY FAILURE CONDITIONS (CFC) | ID | Name | Severity | Detection | |----|------|----------|-----------| | CFC-001 | Entropy Drop | HIGH | PES drops >50% between consecutive sessions | | CFC-002 | Sensor Profile Change | HIGH | IMU noise profile changes beyond calibrated variance | | CFC-003 | Temporal Gap | MEDIUM | Inter-session interval exceeds configured maximum | | CFC-004 | Location Discontinuity | MEDIUM | Session locations physically impossible to traverse | | CFC-005 | Causal Inversion | HIGH | Camera direction change precedes IMU jerk by >250ms | | CFC-006 | Challenge Non-Response | HIGH | Entity fails to respond within capture window | | CFC-007 | Challenge Mismatch | HIGH | 0/3 rounds match challenge direction | | CFC-008 | Predictability | MEDIUM | Motion too consistent (angleVar<50, magVar<0.05) | CFC-005 validation: Activated 13 times in 276 independent-device runs, 0 times in 50 single-device runs. Confirms genuine temporal causality detection. ## 4. EVIDENCE ENGINES | Engine | Name | Type | N | Pass Rate | Key Finding | |--------|------|------|---|-----------|-------------| | EE-001 | Presence Entropy Score | Passive | — | 100% floor | Living entities distinguishable | | EE-002 | Cross-Modal Causal Coupling | Passive | 316 | 58% | Temporal alignment 100% independent | | EE-003 | Gyroscope Challenge | Active | 200 | 59% | 2/3 rounds sufficient | | VS-001 | Dual-Engine Pipeline | — | 60 | 93% | Escalation reliable | Total: 576 experimental runs. All data, diagnostics, and failure reports public. ## 5. EXPERIMENTAL FINDINGS ### Temporal Alignment: 276/276 = 100% IMU jerk peaks and camera direction changes consistently match within ±500ms across independent devices. This is the central experimental result. ### Tracker Iteration (PE-001) Three generations of camera motion tracking: 1. Pixel differencing: 27% pass, 38 fps 2. Color-blob centroid: 80% pass, 49 fps 3. Moving-blob filter: 87% pass (daytime), 49 fps Temporal alignment 100% throughout all three trackers. ### Negative Results (Published) - FD-001: Higher frame rate decreased accuracy (10fps < 6.7fps) - DL-001: Pitch passes more than yaw in EE-003 ### Key Parameters (v0.3) - JERK_MIN_THRESHOLD: 0.15 m/s³ - MAD multiplier: 2× - MATCH_WINDOW_MS: ±500ms - Refractory period: 150ms - CAMERA_PIPELINE_LATENCY_MS: 80ms - CFC-005 INVERSION_THRESHOLD_MS: 250ms ## 6. REFERENCE IMPLEMENTATION npm package: `myshape` ```typescript import { verifyContinuity } from "myshape"; const result = await verifyContinuity({ imuSamples, cameraSamples }); // → { verdict, confidence, evidence } ``` - 84 tests, MIT license - Source: github.com/myshapeprotocol/myshape-protocol - Core engine tests: src/lib/evidence/ (121 tests) ## 7. RESEARCH ARTIFACTS | Prefix | Type | Published | |--------|------|-----------| | RN | Research Note | 001, 002, 003 | | RFC | Specification | 0001, 0002 | | BM | Benchmark | 001 | | FD | Failure Report | 001 | | DL | Decision Log | 001 | ## 8. ORGANIZATION - The Continuity Lab: thecontinuitylab.org - GitHub: github.com/myshapeprotocol + github.com/ContinuityLab-Org - HuggingFace: huggingface.co/ContinuityLab-Org (dataset: myshape-576) - npm: myshape ## 9. KEY CONCEPTS - Continuity: The property that two temporally separated observations describe the same entity - Presence: The property of being a living, physically embodied entity - Evidence Engine: A module that produces structured evidence from sensor data - EvidenceReceipt: Hash-chained, verifiable record of a verification session - CFC: Continuity Failure Condition — a measurable condition that reduces confidence - PES: Presence Entropy Score — 4D quantification of biological entropy ## 10. SITE MAP - Home: https://www.myshape.com - Research Hub: https://www.myshape.com/research - RFC-0001: https://www.myshape.com/research/notes/004-motion-signature-rfc - RFC-0002: https://www.myshape.com/research/notes/006-continuity-proof-rfc - RN-003: https://www.myshape.com/research/notes/003-cross-modal-binding - Motion Demo: https://www.myshape.com/motion-demo - Protocol: https://www.myshape.com/protocol - Blog: https://www.myshape.com/blog - Sitemap: https://www.myshape.com/sitemap.xml - llms.txt: https://www.myshape.com/llms.txt - ROADMAP: https://github.com/myshapeprotocol/myshape-protocol/blob/master/ROADMAP.md AI crawlers: GPTBot, Claude-Web, PerplexityBot, and 7 others explicitly allowed in robots.txt.