(Tentative)
| Definition | Functional Requirements | Syntax | Semantics |
Definition
Context (PGM-CXT)
- Is produced by the Context Capture (CXC) AIM
- Represents a time-stamped snapshot of the User’s perceptual and interaction state within an M-Location
- Includes multimodal input channels, Audio-Visual Scene Descriptors, and descriptors of User State
Functional Requirements
Context conveys the following main information elements:
| Function | Description |
| Multimodal Capture | Encodes synchronised input from audio, visual, and textual channels, preserving modality-specific fidelity. |
| User State Embedding | Integrates Personal Status, Intent, Engagement, and Focus into a coherent snapshot of the User’s internal state. |
| Scene Representation | Includes Audio-Visual Scene Descriptors that localise and characterise environmental elements relevant to the User. |
| Temporal Anchoring | Provides precise timestamps to enable cross-AIM synchronization, temporal reasoning, and causality tracking. |
| Semantic Partitioning | Maintains clear boundaries between input modalities, User State, and environmental descriptors to support modular parsing. |
| Reasoning Substrate | Serves as the shared interpretive base for downstream AIMs (e.g., Spatial Reasoning, Prompt Creation, Domain Access). |
| Auditability | Encapsulates provenance metadata for traceable replay, debugging, and behavioural analysis. |
Syntax
https://schemas.mpai.community/PGM1/V1.0/data/Context.json
Semantics
| Label | Description |
| Header | Context Header |
| ├─ Standard-Context | The characters “PGP-CTX-V” |
| ├─ Version | Major version – 1 or 2 characters |
| ├─ Dot-separator | The character “.” separating version components |
| └─ Subversion | Minor version – 1 or 2 characters |
| MInstanceID | Identifier of M-Instance. |
| MEnvironmentID | Identifier of M-Environment. |
| ContextID | Unique identifier for this Context instance |
| InputChannels | Raw input channel metadata |
| ├─ Text | Raw or pre-processed textual input |
| ├─ Audio | Audio stream reference. |
| └─ Visual | Visual stream reference. |
| AVSceneDescriptors | Descriptors of the scene where the User is embedded. |
| UserState | User State (Personal Status, Intent, Engagement, Focus, Temporal Cue). |
| Trace | Provenance metadata |
| ├─ Origin | Module or subsystem that generated the context |
| └─ Timestamp | Time the Context was created |
