| name | super-ai-ml-foundation |
| description | AI/ML foundations: model selection, prompt design, RAG, embeddings, and vector search. Use for core AI app design and build. |
Super AI/ML Foundation
Overview
Establish the core AI/ML architecture before building agents or ops layers.
User Intent Examples
- "Need help with AI Engineering for my product/site."
- "Create a plan for Prompt Engineering."
- "Audit or improve RAG Engineering."
Workflow
- Confirm use case, success criteria, latency, and cost targets.
- Select model family and deployment approach (hosted vs self-hosted).
- Design prompts, tool interfaces, and guardrails.
- Plan and implement RAG: chunking, embeddings, indexing, retrieval.
- Validate relevance, quality, and failure modes with small tests.
- Document tradeoffs, risks, and next experiments.
Minimal Intake Questions
- Primary goal or outcome
- Scope (pages, systems, teams, or timeframe)
- Constraints (tools, budget, timeline)
Output Format
- Use-case brief and success metrics
- Model choice with rationale
- Prompt and tool plan
- RAG architecture plan
- Risks, mitigations, and next steps
Routing Map (Modules)
- AI Engineering ->
references/modules/ai-engineer.md
- Prompt Engineering ->
references/modules/prompt-engineering-patterns.md
- RAG Engineering ->
references/modules/rag-engineer.md
Bundled References
references/modules/
scripts/
assets/
agents/
Compatibility Notes
- If any module references slash commands or tool-specific paths, translate them into plain-language steps.
- Keep outputs platform-agnostic unless the user specifies a specific tool, stack, or agent.
Guardrails
- Do not claim benchmark results without data.
- Separate measured results from hypotheses.
- Prefer the smallest viable model and simplest retrieval setup.