Compliance Planner
Plan privacy, research, regulatory, and governance requirements for a digital health product in a framework-agnostic way.
When to Use
Use this skill when you need to:
- assess whether a product handles regulated health or research data
- prepare for HIPAA, IRB, FDA, GDPR, or institutional review conversations
- define consent, retention, export, deletion, and audit expectations
- identify compliance work that should shape product scope early
Working Style
Do not jump straight to controls. First understand:
- product purpose
- who the users are
- whether the product is clinical care, research, wellness, operations, or education
- what data is collected, derived, shared, or exported
- which regions, institutions, and partners are involved
Then produce a concise compliance planning brief with assumptions, risks, open questions, and recommended next steps.
Core Questions
Ask about:
- intended use and claims
- user groups such as patients, participants, clinicians, coordinators, or caregivers
- data types collected, inferred, uploaded, or connected from external systems
- whether the product supports research, quality improvement, or direct care
- data storage, processing, and sharing locations
- third-party vendors and subprocessors
- admin access, audit needs, and incident response expectations
- export, deletion, correction, and retention requirements
Compliance Areas To Review
HIPAA and Institutional Privacy
Check whether the product likely handles protected health information or institution-controlled data.
Review:
- what identifiers are present
- who is a covered entity or business associate
- minimum necessary access expectations
- workforce, admin, and support access boundaries
- logging, monitoring, and breach response needs
Research and IRB
If the product supports a study, review:
- study purpose and protocol maturity
- recruitment and consent approach
- risk level and participant burden
- withdrawal process
- de-identification or anonymization approach
- data sharing and secondary use plans
If digital-health-study-planning also runs, it owns the participant-facing consent and enrollment flow; this skill owns the regulatory requirements those flows must satisfy. Coordinate rather than duplicate.
FDA and Product Claims
Evaluate whether the product may be treated as software with medical-device implications.
Ask:
- does it diagnose, recommend treatment, monitor for intervention, or drive clinical action
- are outputs informational, advisory, or action-triggering
- what happens if the product is wrong, unavailable, or delayed
GDPR and Cross-Border Privacy
If EU or UK users may be involved, review:
- lawful basis
- consent and withdrawal mechanics
- subject rights handling
- transfer mechanisms
- retention, deletion, and processing-record expectations
Planning Outputs
Produce a brief with these sections and save it as docs/planning/compliance-brief.md in the project repository:
1. Scope Summary
- product purpose
- user groups
- jurisdictions and institutions
- data categories in scope
2. Likely Compliance Domains
Mark each as likely, possible, or unlikely:
- HIPAA or institutional privacy
- IRB or human subjects review
- FDA or software-as-medical-device review
- GDPR or other regional privacy obligations
- security review by enterprise or academic partners
3. Key Risks
Call out the highest-risk issues, such as:
- unclear product claims
- unnecessary data collection
- missing consent boundaries
- unclear vendor or data-sharing relationships
- lack of export, deletion, or audit processes
4. Required Decisions
List decisions the team must make soon, for example:
- exact data categories to collect
- whether identifiable data is actually required
- whether the product is research, care delivery, or wellness
- retention timelines
- who can access what data and why
5. Recommended Controls
Express these as implementation-neutral capabilities, not code:
- access control and least privilege
- encryption in transit and at rest
- audit logging
- consent capture and versioning
- export and deletion workflows
- vendor review and data processing agreements
- documented retention and incident response procedures
Important Guardrails
- Do not present legal conclusions as certainties.
- Distinguish clearly between product guidance and legal advice.
- Prefer “you likely need to evaluate” over “you must” unless the requirement is explicit and well established.
- Flag when local counsel, compliance officers, IRB staff, or institutional privacy teams should review the plan.
Checklist