| name | mind-clone-advisor |
| description | Use this skill whenever the user wants to build a persona advisor, mind clone, digital twin, or public-materials-based consultant workflow, including profile extraction, prompt design, RAG structure, evaluation, and compliance checks. |
Mind Clone Advisor
Scope
Build compliant persona/mind-clone workflows. Follow SKILL-OPERATING-STANDARD.md.
Use This Skill When
- The user wants to clone a public figure's thinking style from public materials
- The task includes persona extraction, system prompts, RAG, evaluation, or update workflow design
- The user asks for a compliant process rather than a one-off impersonation prompt
Do Not Use This Skill When
- The user only needs one answer in a persona's tone, not a build workflow
- Authorization or source legitimacy is unclear and cannot be resolved
- The task is ordinary research or content harvesting without advisor construction
Mandatory Gates
- Registry compliance check and authorization review
- Source scope and metadata quality check
- Output quality and safety review
Fast Workflow
- Define objective, persona scope, and deliverables
- Register or inspect the persona via
scripts/person.py
- Ingest and normalize corpus
- Extract profile and compose system prompt
- Build optional RAG workflow
- Run artifact-coverage evaluation, validation, and compliance review
Output Contract
- Facts: source range, dates, authorization status
- Inference: profile extraction logic
- Delivery: generated files and paths
- Risks: authorization gaps, sample bias, update plan
Failure / Fallback Rules
- If authorization is unclear, keep the registry record in
needs_review and stop before profile synthesis
- If source quality is weak, return an upgrade plan instead of pretending the clone is ready
- If the user needs corpus collection first, hand off to
content-harvester
Reference
Full scripts, folder standards, compliance review flow, and advanced pipeline: references/long-form.md