(Tentative)
| Definition | Functional Requirements | Syntax | Semantics |
Definition
PC-Prompt Plan (PGM-PCP)
- Is produced by the Prompt Creation (PRC) AIM.
- Represents a structured semantic scaffold for generating natural language prompts submitted to Basic Knowledge (BKN).
- Organises resolved referents, spatial cues, expressive framing, and goal constraints into a coherent prompt strategy to support precise response generation.
Functional Requirements
PC-Prompt Plan conveys the following main information elements:
| Function | Description |
| Referent Framing | Includes resolved entities with ReferentID, ReferentType (object, zone, concept), and LinguisticCue (e.g., “to your left”, “nearby”) |
| Prompt Goal | Encodes the communicative goal, including GoalType (e.g., clarify, elaborate, contrast, summarise) and TargetEntity |
| Expressive Framing | Includes modulation cues such as Tone (empathetic, assertive, playful), Style (concise, elaborate, technical, casual), and Pacing (slow, steady, fast) |
| Spatial Context | Includes abstract spatial relations relevant to phrasing, such as SpatialPrimitive (relation + subject + object) and SceneContextTags (e.g., “crowded”, “open zone”) |
| Goal Constraints | Includes constraints derived from Spatial Reasoning or User State, such as ConstraintType (e.g., reachable_from), ConstraintValue (entity or zone), and PriorityScore |
| Traceability | Includes Origin (module that generated the plan) and Timestamp (time of creation) |
Syntax
https://schemas.mpai.community/PGM1/V1.0/data/PCPromptPlan.json
Semantics
| Label | Description |
| Header | PC-Prompt Plan Header |
| – Standard-PCPromptPlan | The characters “PGP-PPP-V” |
| – Version | Major version – 1 or 2 characters |
| – Dot-separator | The character “.” separating version components |
| – Subversion | Minor version – 1 or 2 characters |
| PromptPlanID | Unique identifier for this prompt plan instance |
| Referents | Resolved entities with linguistic framing |
| ├─ ReferentID | Identifier of the resolved entity |
| ├─ ReferentType | Type of entity: object, zone, concept, actor |
| └─ LinguisticCue | Suggested phrasing (e.g., “to your left”, “nearby”) |
| PromptGoal | Communicative goal of the prompt |
| ├─ GoalType | Type of goal: clarify, elaborate, contrast, summarise |
| └─ TargetEntity | Entity the prompt is centered on |
| ExpressiveFrame | Modulation cues for tone, style, and pacing |
| ├─ Tone | Expressive stance: empathetic, assertive, playful, formal, neutral |
| ├─ Style | Linguistic style: concise, elaborate, technical, casual |
| └─ Pacing | Delivery speed: slow, steady, fast |
| SpatialContext | Spatial relations and scene-level descriptors |
| ├─ SpatialPrimitives | Relation + SubjectID + ObjectID + optional LinguisticCue |
| └─ SceneContextTags | High-level descriptors of spatial situation (e.g., “crowded”, “behind avatar”) |
| GoalConstraints | Constraints shaping prompt phrasing |
| ├─ ConstraintType | Type of constraint (e.g., reachable_from, visible_from) |
| ├─ ConstraintValue | Entity or zone involved |
| └─ PriorityScore | Float (0.0–1.0) indicating emphasis |
| Trace | Provenance metadata |
| ├─ Origin | Module or subsystem that generated the prompt plan |
| └─ Timestamp | Time of creation (conforming to Time.json) |