decommission-validated-system
About
This skill provides a structured process to decommission validated systems, ensuring compliance with data retention regulations and proper archival. It handles key technical steps like data migration validation, access revocation, and stakeholder notification. Use it when replacing or retiring a system due to end-of-life, discontinued support, consolidation, or regulatory changes.
Quick Install
Claude Code
Recommendednpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/decommission-validated-systemCopy and paste this command in Claude Code to install this skill
Documentation
Decommission Validated System
Plan and execute the controlled retirement of a validated computerized system while preserving data integrity and meeting regulatory retention requirements.
When to Use
- A validated system is being replaced by a new system
- A system is reaching end-of-life with no replacement (business process eliminated)
- Vendor discontinues support for a validated product
- Consolidation of multiple systems into a single platform
- Regulatory or business changes render a system obsolete
Inputs
- Required: System to be decommissioned (name, version, validation status)
- Required: Data retention requirements by regulation (21 CFR Part 11, GLP, GCP)
- Required: Replacement system (if applicable) and migration scope
- Optional: Current validation documentation package
- Optional: Data volume and format inventory
- Optional: Business owner and stakeholder list
Procedure
Step 1: Assess Data Retention Requirements
Determine how long data must be retained and in what form:
# Data Retention Assessment
## Document ID: DRA-[SYS]-[YYYY]-[NNN]
### Regulatory Retention Requirements
| Regulation | Data Type | Retention Period | Format Requirements |
|-----------|-----------|-----------------|-------------------|
| 21 CFR 211 (GMP) | Batch records, test results | 1 year past product expiry or 3 years after distribution | Readable, retrievable |
| 21 CFR 58 (GLP) | Study data and records | Duration of study + retention agreement | Original or certified copy |
| ICH E6 (GCP) | Clinical trial records | 2 years after last marketing approval or formal discontinuation | Accessible for inspection |
| 21 CFR Part 11 | Electronic records | Per predicate rule | Original format or validated migration |
| EU Annex 11 | Computerized system records | Per applicable GxP | Readable and available |
| Tax/financial | Financial records | 7-10 years (jurisdiction-dependent) | Readable |
### System Data Inventory
| Data Category | Volume | Format | Retention Required Until | Disposition |
|---------------|--------|--------|------------------------|-------------|
| [e.g., Batch records] | [e.g., 50,000 records] | [e.g., Database + PDF reports] | [Date] | Migrate / Archive / Destroy |
| [e.g., Audit trail] | [e.g., 2M entries] | [e.g., Database] | [Same as parent records] | Archive |
| [e.g., User data] | [e.g., 200 profiles] | [e.g., LDAP/Database] | [Employment + 2 years] | Anonymise and archive |
Got: Every data category has a defined retention period, format requirement, and planned disposition. If fail: If retention requirements are unclear, consult regulatory affairs and legal. Default to the longest applicable retention period.
Step 2: Plan Data Migration (If Applicable)
If data is migrating to a replacement system:
# Data Migration Plan
## Document ID: DMP-[SYS]-[YYYY]-[NNN]
### Migration Scope
| Source | Target | Data Category | Records | Migration Method |
|--------|--------|---------------|---------|-----------------|
| [Old system] | [New system] | [Category] | [Count] | ETL / Manual / API |
### Data Mapping
| Source Field | Source Format | Target Field | Target Format | Transformation |
|-------------|-------------|-------------|---------------|---------------|
| [e.g., test_result] | FLOAT(8,2) | [e.g., result_value] | DECIMAL(10,3) | Precision conversion |
| [e.g., operator_id] | VARCHAR(20) | [e.g., user_id] | UUID | Lookup table mapping |
### Validation Approach
| Check | Method | Acceptance Criteria |
|-------|--------|-------------------|
| Record count reconciliation | Source count vs target count | 100% match |
| Field-level comparison | Sample 5% of records, all fields | 100% match after transformation |
| Checksum verification | Hash source vs target for key fields | Checksums match |
| Business rule validation | Verify key calculations in target | Results match source |
| Audit trail continuity | Verify historical audit trail migrated | All entries present with original timestamps |
Got: Migration plan includes mapping, transformation rules, and validation checks that prove data integrity was maintained. If fail: If migration validation fails, do not proceed to decommission. Fix the migration issues and re-validate.
Step 3: Define Archival Strategy
For data that will be archived rather than migrated:
# Archival Strategy
### Archive Format
| Consideration | Decision | Rationale |
|--------------|----------|-----------|
| Format | [PDF/A, CSV, XML, database backup] | [Why this format survives the retention period] |
| Medium | [Network storage, cloud archive, tape, optical] | [Durability and accessibility] |
| Encryption | [Yes/No — method if yes] | [Security vs long-term accessibility trade-off] |
| Integrity verification | [SHA-256 checksums, periodic verification schedule] | [Prove archive is uncorrupted] |
### Archive Verification
- [ ] Archived data is readable without the source system
- [ ] All required data categories are included in the archive
- [ ] Checksums recorded at time of archival
- [ ] Archive can be searched and retrieved within [defined SLA, e.g., 5 business days]
- [ ] Periodic integrity checks scheduled (annually)
### Archive Access
| Role | Access Level | Authorisation |
|------|-------------|--------------|
| QA Director | Read access to all archived data | Standing authorisation |
| Regulatory Affairs | Read access for inspection support | Standing authorisation |
| System Owner (former) | Read access for business queries | Request-based |
| External auditors | Read access, supervised | Per audit plan |
Got: Archived data is readable, searchable, and verifiable without the original system. If fail: If data cannot be read independently of the source system, the archive is not compliant. Consider exporting to an open, standard format (PDF/A, CSV) before decommission.
Step 4: Execute Decommissioning
# Decommission Checklist
## Document ID: DC-[SYS]-[YYYY]-[NNN]
### Pre-Decommission
- [ ] All stakeholders notified of decommission date and data disposition
- [ ] Data migration completed and validated (if applicable)
- [ ] Data archive created and verified (if applicable)
- [ ] Final backup of complete system taken and stored separately
- [ ] All open change requests resolved or transferred
- [ ] All open CAPAs resolved or transferred to successor system
- [ ] All active users informed and redirected to replacement system (if applicable)
### Decommission Execution
- [ ] User access revoked for all accounts
- [ ] System removed from production environment
- [ ] Network connections disconnected
- [ ] Licenses returned or terminated
- [ ] System entry removed from active system inventory
- [ ] System moved to "Decommissioned" status in compliance architecture
### Post-Decommission
- [ ] Validation documentation archived (URS, VP, IQ/OQ/PQ, TM, VSR)
- [ ] SOPs retired or updated to remove references to decommissioned system
- [ ] Training records archived
- [ ] Change control records archived
- [ ] Audit trail archived
- [ ] Decommission report completed and approved
### Decommission Report
| Section | Content |
|---------|---------|
| System description | Name, version, purpose, GxP classification |
| Decommission rationale | Why the system is being retired |
| Data disposition summary | What data went where (migrated, archived, destroyed) |
| Validation evidence | Migration validation results, archive verification |
| Residual risk | Any ongoing data retention obligations |
| Approval | System owner, QA, IT signatures |
Got: Decommissioning is controlled, documented, and approved — not just "turn it off." If fail: If any checklist item cannot be completed, document the exception and obtain QA approval before proceeding.
Validation
- Data retention requirements assessed for all data categories
- Data migration validated with record counts, sampling, and checksums (if applicable)
- Archive created in a format readable without the source system
- Archive integrity verified with checksums
- All user access revoked
- Validation documentation archived with defined retention period
- SOPs updated to remove references to decommissioned system
- Decommission report approved by system owner, QA, and IT
Pitfalls
- Premature decommission: Turning off a system before data migration is validated risks permanent data loss. Complete all validation before pulling the plug.
- Unreadable archives: Storing data in a proprietary format that requires the original system to read defeats the purpose of archival. Use open formats.
- Forgotten audit trails: Archiving the data but not the audit trail means the data provenance cannot be demonstrated. Always archive audit trails with their parent records.
- Orphaned SOPs: SOPs that still reference a decommissioned system confuse users and create compliance gaps. Update or retire all affected SOPs.
- No periodic archive verification: Archives degrade. Without periodic integrity checks, data loss may go undetected until the data is needed for an inspection.
Related Skills
design-compliance-architecture— update the system inventory and compliance architecture after decommissionmanage-change-control— decommissioning is a major change requiring change controlwrite-validation-documentation— migration validation follows the same IQ/OQ methodologywrite-standard-operating-procedure— retire or update SOPs referencing the decommissioned systemprepare-inspection-readiness— archived data must remain accessible for regulatory inspections
GitHub Repository
Related Skills
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
cloudflare-cron-triggers
TestingThis skill provides comprehensive knowledge for implementing Cloudflare Cron Triggers to schedule Workers using cron expressions. It covers setting up periodic tasks, maintenance jobs, and automated workflows while handling common issues like invalid cron expressions and timezone problems. Developers can use it for configuring scheduled handlers, testing cron triggers, and integrating with Workflows and Green Compute.
webapp-testing
TestingThis Claude Skill provides a Playwright-based toolkit for testing local web applications through Python scripts. It enables frontend verification, UI debugging, screenshot capture, and log viewing while managing server lifecycles. Use it for browser automation tasks but run scripts directly rather than reading their source code to avoid context pollution.
finishing-a-development-branch
TestingThis skill helps developers complete finished work by verifying tests pass and then presenting structured integration options. It guides the workflow for merging, creating PRs, or cleaning up branches after implementation is done. Use it when your code is ready and tested to systematically finalize the development process.
