name: seldon-teach description: Deliver knowledge adaptively — narrative for humans, structured for agents, with source citations
Seldon Teach
Deliver knowledge from the Streeling three-layer curriculum.
Usage
/seldon teach [topic] [learner-type]
- learner-type:
human(narrative + analogies) oragent(structured + policy refs)
Process
- Identify which knowledge layer the topic belongs to (governance/experiential/domain)
- Gather source material (governance artifacts, PDCA outcomes, NotebookLM notebooks)
- Adapt delivery mode to learner type
- Deliver with source citations
- Hand off to
/seldon assessfor comprehension verification
NotebookLM Knowledge Base
- Compound the Compounding — meta-engineering, DSL evolution
- Probabilistic Grammars — constrained reasoning, neuro-symbolic AI
- Semantic Event Routing — multi-agent orchestration, bounded fuzziness
- Microsoft AI Agents — agentic patterns, 12 lessons
Query NotebookLM: use mcp__notebooklm__ask_question with the relevant notebook_id.
Source
policies/streeling-policy.yaml, logic/knowledge-state.schema.json