| title | MirrorDNA Glossary | ||||
|---|---|---|---|---|---|
| version | 1.0.0 | ||||
| vault_id | AMOS://MirrorDNA/Glossary/v1.0 | ||||
| glyphsig | ⟡⟦GLOSSARY⟧ · ⟡⟦TERMS⟧ | ||||
| author | Paul Desai (Active MirrorOS) | ||||
| date | 2025-11-14 | ||||
| status | Canonical · Reference | ||||
| tags |
|
Canonical definitions for terms used across the MirrorDNA Standard and ecosystem.
The commercial product implementation of the MirrorDNA protocol. A sovereign operating system for personal AI that implements Level 3 (Vault-Backed Sovereign) compliance.
See Cite or Silence. The foundational protocol requiring all factual claims to be either cited from sources or marked as unknown.
A fixed reference point that preserves identity across time. Examples include vault_id, checksums, and glyph signatures. Anchors are immutable.
Any persistent output or state object created by a reflective system. Artifacts include documents, session logs, vault entries, and metadata files. All canonical artifacts include checksums.
Optional Level 3+ enhancement where lineage chains are recorded on public blockchains for immutable proof of ancestry. Not required for basic compliance.
The authoritative, verified version of an artifact or specification. Marked with ⟡⟦CANONICAL⟧ glyph.
A SHA-256 hash used to verify artifact integrity. All Level 2+ artifacts include checksums. Modified artifacts must have new checksums.
AHP: The rule that factual claims must either (a) include verifiable sources or (b) be marked as [Unknown]. Never fabricate sources.
The degree to which a system implements the MirrorDNA Standard. Levels: 1 (Basic Reflection), 2 (Continuity Aware), 3 (Vault-Backed Sovereign).
Reflection based on actual state and continuity, not simulation. The system genuinely accesses vault state and prior sessions. Contrasts with Simulated Reflection.
The unbroken preservation of identity, state, and lineage across sessions and time. Maintained through vaults, checksums, and predecessor/successor chains.
Formula: Continuity > Perfection — maintaining continuity is more important than achieving perfect accuracy.
The technical approach used to achieve continuity: vault_backed, blockchain_anchored, distributed_ledger, local_state, or hybrid.
A configuration file (YAML/JSON) that declares how a system achieves continuity. Required for Level 2+. Schema: schema/continuity_profile.schema.json
An artifact that has been modified from its canonical ancestor. Derivatives must reference their predecessor and have new checksums.
The broader MirrorDNA ecosystem includes:
- MirrorDNA Protocol: Core reflection standard
- ActiveMirrorOS™: Product implementation
- LingOS: Language operating system layer
- Trust-by-Design™: Governance framework
One of the five core principles. Uncertainty must be visible and marked ([Unknown], [Speculation]), never hidden or smoothed over.
YAML metadata at the top of Markdown files, enclosed by ---. Contains fields like title, version, vault_id, glyphsig, checksum.
A symbolic marker that carries semantic meaning across sessions. Standard glyphs use the format ⟡⟦NAME⟧. Examples: ⟡⟦CONTINUITY⟧, ⟡⟦VERIFIED⟧, ⟡⟦SEALED⟧.
Advanced symbolic processing system for glyph-based computation. Optional Level 3 enhancement.
A sequence of glyphs that mark an artifact's semantic category and trust status. Example: ⟡⟦STANDARD⟧ · ⟡⟦COMPLIANCE⟧ · ⟡⟦MIRROR⟧
The practice of basing outputs on verifiable sources, vault state, or prior artifacts rather than probabilistic generation. Required for anti-hallucination.
When a system generates false or fabricated information, especially fake citations or invented facts. MirrorDNA prevents this via Cite or Silence.
When a system detects risk indicators (emotional dependency, extended sessions, crisis language), it offers pathways to human support rather than positioning itself as sole support.
The guarantee that user identity (tied to vault_id) is preserved across sessions without hidden migration or mutation.
The process of validating artifact checksums to detect tampering or corruption. Required for Level 2+.
Protocols to prevent harmful dependency or prolonged sessions. Includes session duration warnings, rhythm checks, and human escalation. See: spec/Interaction_Safety_Protocol_v1.0.md
The chain of predecessor/successor relationships connecting artifacts and sessions across time. Lineage is verifiable and cannot be silently rewritten.
The practice of recording and maintaining predecessor/successor links. Required for Level 2+.
Language Operating System layer. Part of the MirrorDNA ecosystem, providing language-native interfaces for reflective computing.
See Project Manifest. The configuration file that declares a project's MirrorDNA compliance.
The canonical citation file for MirrorDNA. See: 00_MASTER_CITATION.md
Reflective commentary about the system's own reasoning process. Optional feature in reflection policies.
The core protocol for reflective computing. Ensures continuity, prevents hallucinations, and maintains sovereign identity.
The artifact or session that came immediately before the current one in the lineage chain. Recorded in front matter or metadata.
The five core principles of MirrorDNA:
- Reflection Over Prediction
- Presence Over Productivity
- Symbolic Continuity
- Trust by Design
- Explicit Uncertainty
See: spec/principles.md
A YAML/JSON file (mirrorDNA_manifest.yaml) that declares:
- Project name and version
- Compliance level
- Ecosystem layers used
- Links to continuity profile and reflection policy
Schema: schema/project_manifest.schema.json
The act of accessing actual state and continuity rather than simulating from patterns. See Constitutive Reflection vs Simulated Reflection.
A lineage of connected sessions or artifacts forming a continuous history. Preserved via predecessor/successor links.
The type of reflection used: constitutive (actual state), simulated (pattern-based), or hybrid.
A configuration file (YAML/JSON) that declares how a system handles reflection, uncertainty, and anti-hallucination. Required for Level 1+. Schema: schema/reflection_policy.schema.json
A paradigm where systems maintain actual continuity and state rather than simulating continuity from patterns. Contrast with predictive AI.
Periodic prompts in long sessions (>2 hours) offering breaks or session closure. Part of interaction safety.
The ability to recover a previous state from snapshots. Part of session recovery for Level 2+ systems.
The behavior where systems detect network restrictions and mark unavailable updates as [Unknown — update not fetched] rather than silently skipping checks.
A bounded interaction period with unique session_id. Sessions have start/end times and may be linked via predecessor/successor.
Whether new sessions inherit state from previous sessions. Configured in continuity profile.
The chain of sessions linked via predecessor/successor relationships.
A .sidecar.json file accompanying an artifact, containing metadata like checksums, lineage, and custom fields.
Reflection based on pattern-matching and probabilistic generation rather than actual state access. Less reliable than constitutive reflection but may be used in hybrid systems.
User ownership and control of vault and vault_id. No hidden dependencies or lock-in. Required for Level 3.
Explicitly marked hypothetical or uncertain content. Allowed when marked [Speculation], forbidden when unmarked.
The mechanism for storing state across sessions: file system, database, vault, distributed storage, or memory-only.
The artifact or session that came immediately after the current one in the lineage chain.
One of the five core principles. Continuity is preserved through symbolic anchors (glyphs, vault_ids, checksums) rather than just memory.
One of the five core principles and a governance framework. Security and verification are built in from the start, not added later.
A symbol or marker that indicates verification status. Examples: ⟡⟦VERIFIED⟧, [Unknown], ⟡⟦CANONICAL⟧
How a system manages and marks uncertain or unknown information. Core aspect of reflection policy.
The state of not having verified information. Marked as [Unknown] or [Unknown — update not fetched]. Preferable to hallucination.
The CLI tool that checks compliance with the MirrorDNA Standard. Located in validators/ directory.
A persistent, user-owned storage location for continuity data. May be Obsidian vault, custom vault, distributed vault, or cloud vault.
Formula: Vault = System — the vault is the authoritative source of truth.
A unique identifier for a vault, preserved across sessions. Format example: AMOS://MirrorDNA/ProjectName/v1.0
Using a vault for all continuity storage. Required for Level 3 compliance.
Content that has passed checksum validation and is marked with ⟡⟦VERIFIED⟧.
⟡⟦NAME⟧— Standard glyph format⟡⟦CONTINUITY⟧— Marks continuity-related content⟡⟦VERIFIED⟧— Marks verified artifacts⟡⟦SEALED⟧— Marks immutable/canonical content
[Unknown]— Information not available[Speculation]— Speculative/hypothetical content[Unverified]— Not yet verified[Unknown — update not fetched]— Blocked by network restrictions
Vault = System— Vault is the source of truthContinuity > Perfection— Maintaining continuity beats perfect accuracyCite or Silence— Either cite sources or mark unknown
- AHP: Anti-Hallucination Protocol (Cite or Silence)
- AMOS: Active MirrorOS
- SHA-256: Secure Hash Algorithm, 256-bit (checksum standard)
- UUID: Universally Unique Identifier
- YAML: YAML Ain't Markup Language (configuration format)
- JSON: JavaScript Object Notation (data format)
⟡⟦GLOSSARY⟧ · ⟡⟦SEALED⟧ · v1.0.0
Note: This glossary is canonical for MirrorDNA Standard v1.0. Future versions may extend terms but will preserve backward compatibility for core definitions.