(Tentative)

Function Reference Model Input/Output Data
SubAIMs JSON Metadata Profiles

Function

The Basic Knowledge AIM (PGM-BKN) is a semantic engine that receives structured prompts from Prompt Creation (PRC), Domain Access (DAC), State Adaptation (SAD), and Personality Alignment (PAL). Its primary function is to transform these prompts into progressively refined responses by applying domain-specific knowledge, contextual reasoning, and expressive modulation – not necessarily relying on a general-purpose language model.

Internally, PGM-BKN may perform the following operations:

  • Prompt Interpretation: Parses structured prompts to extract communicative intent, domain references, and expressive framing.
  • Knowledge Retrieval: Accesses internal knowledge structures – such as ontologies, rule sets, or curated response templates – to identify relevant content.
  • Contextual Reasoning: Integrates User State, interaction history, and domain constraints to adapt responses to the current situation.
  • Response Refinement: Iteratively improves candidate responses by aligning them with inferred goals, Personality modulation, and interaction tone.
  • Trace Construction: Generates a BK Response Trace containing the response, its derivation path, and any knowledge sources or transformations applied.
  • Output Dispatch: Sends the refined response trace to A-User Control for orchestration, rendering, or further modulation.

The resulting outputs enable A-User Control to deliver responses that are meaningfully structured and knowledge-aware, adapted to current interaction state, and aligned with the User’s expressive profile and interaction goals.

Reference Model

Figure 7 gives the Reference Model of Basic Knowledge (PGM-BKN).

Figure 7 – The Reference Model of Basic Knowledge (PGM-BKN)

Input/Output Data

Table 16 gives Input and Output Data of Basic Knowledge (PGM-BKN). All data are Text.

Table 16 – Input and Output Data of Basic Knowledge (PGM-BKN)

Input Description
PC-Prompt Prompt by Prompt Creation
DA-Prompt Prompt by Domain Access
UR-Prompt Prompt by State Adaptation
PA-Prompt Prompt by Personality Alignment
BK Response Trace Metadata for Enriched Response and audit traceability from PGM-AUC.
Output Description
Initial Response Response to PC-Prompt
Enhanced Response Response to DA-Prompt
Refined Response Response to SA-Prompt
Final Response Response to PA-Prompt
BK Query Directive Trigger to retrieve knowledge or shape response from PGM-AUC.

SubAIMs

No SubAIMs.

Table 2 maps BKN Inputs/Outputs to Unified Messages

Table 2 — BKN Inputs/Outputs mapped to Unified Messages

BKN Data Name Role Origin / Destination Unified Schema Mapping
PC-Prompt Input From PRC Input → Payload (prompt text / multimodal). Correlate with Envelope.CorrelationId. MUST carry Trace.Origin and Trace.Timestamp from PRC.
Initial Response Output To DAC Status → Result (first LLM output for domain refinement). Maintain Envelope.CorrelationId; includes Trace.Origin=BKN and Trace.Timestamp.
DA-Prompt Input From DAC Input → Payload addendum (domain constraints, priorities, vocabulary). Bound to same Envelope.CorrelationId; DAC adds Trace.Origin/Trace.Timestamp.
Enhanced Response Output To USR Status → Result (domain-refined response). Maintain Envelope.CorrelationId; includes Trace.Origin=BKN and Trace.Timestamp.
US-Prompt Input From USR Input → Payload addendum (tone, persona, formatting). Correlate via Envelope.CorrelationId; include Trace.Origin/Trace.Timestamp.
Refine Response Output To PAL Status → Result (style-refined response). Maintain Envelope.CorrelationId; includes Trace.Origin=BKN and Trace.Timestamp.
PA-Prompt Input From PAL Input → Payload addendum (task plan, tool-use intent). Correlated via Envelope.CorrelationId; include Trace.Origin/Trace.Timestamp.
Final Response Output To AUF Status → Result (final response for actuation). Maintain Envelope.CorrelationId; includes Trace.Origin=BKN and Trace.Timestamp.

✅ Why this works:

JSON Metadata

https://schemas.mpai.community/PGM1/V1.0/AIMs/BasicKnowledge.json

Profiles

No Profiles.