arckit-cli 6.1.1__py3-none-any.whl

arckit_cli-6.1.1.data/data/share/arckit/.arckit/templates/uk-gov-atrs-template.md

@@ -0,0 +1,1087 @@
+# Algorithmic Transparency Recording Standard (ATRS)
+
+> **Template Origin**: Official | **ArcKit Version**: [VERSION] | **Command**: `/arckit:atrs`
+
+## Document Control
+
+<!-- DOC-CONTROL-HEADER -->
+<!-- Resolved at command-execution time to _partials/document-control-uk.md or _partials/document-control-uae.md based on plugin userConfig classification_scheme + governance_framework. See _partials/RENDERING.md (when present). -->
+
+## Revision History
+
+| Version | Date | Author | Changes | Approved By | Approval Date |
+|---------|------|--------|---------|-------------|---------------|
+| [VERSION] | [DATE] | ArcKit AI | Initial creation from `/arckit.[COMMAND]` command | [PENDING] | [PENDING] |
+
+## Document Purpose
+
+[Brief description of what this document is for and how it will be used]
+
+---
+
+## About This Template
+
+This template follows the UK Government's Algorithmic Transparency Recording Standard (ATRS), which is **MANDATORY** for:
+
+- All central government departments
+- Arm's length bodies (ALBs) delivering public/frontline services
+- Organizations directly interacting with the general public
+
+**Purpose**: The ATRS helps public sector organizations provide clear information about how and why they use algorithmic tools, including AI systems.
+
+**Structure**:
+
+- **Tier 1**: Summary information for the general public (clear, simple language)
+- **Tier 2**: Detailed technical information for specialists, journalists, researchers
+
+**Resources**:
+
+- ATRS Guidance: https://www.gov.uk/government/publications/guidance-for-organisations-using-the-algorithmic-transparency-recording-standard
+- ATRS Template: https://www.gov.uk/government/publications/algorithmic-transparency-template
+- Contact: algorithmic-transparency@dsit.gov.uk
+
+---
+
+# TIER 1: Summary Information
+
+*For the general public - use clear, simple language*
+
+## 1. Basic Information
+
+### 1 - Name
+
+**Tool Name**: [Name of the algorithmic tool]
+
+### 2 - Description
+
+**Brief Description** (1-2 sentences for public understanding):
+[Describe what the tool does in plain English]
+
+### 3 - Website URL
+
+**More Information**: [URL to tool information page]
+
+### 4 - Contact Email
+
+**Contact**: [email@department.gov.uk]
+
+---
+
+## 2. Organization and Phase
+
+### Organization
+
+- **Department/Organization**: [e.g., Department for Work and Pensions]
+- **Organization Type**: [ ] Ministerial Department / [ ] Non-Ministerial Department / [ ] Executive Agency / [ ] Non-Departmental Public Body / [ ] Local Authority / [ ] NHS / [ ] Other
+
+### Function
+
+[ ] Benefits and welfare
+[ ] Crime and policing
+[ ] Education
+[ ] Healthcare
+[ ] Immigration
+[ ] Justice
+[ ] Licensing and regulation
+[ ] Tax and revenue
+[ ] Transport
+[ ] Other: _______________
+
+### Geographic Region
+
+[ ] England
+[ ] Scotland
+[ ] Wales
+[ ] Northern Ireland
+[ ] UK-wide
+[ ] Other: _______________
+
+### Phase
+
+[ ] Pre-deployment (planning/development)
+[ ] Private Beta (limited testing)
+[ ] Public Beta (wider testing)
+[ ] Production (live use)
+[ ] Retired (no longer in use)
+
+**Start Date**: [YYYY-MM or YYYY-MM-DD]
+**End Date** (if retired): [YYYY-MM or YYYY-MM-DD]
+
+---
+
+# TIER 2: Detailed Information
+
+*For specialists, journalists, researchers - more technical detail*
+
+---
+
+## Section 1: Owner and Responsibility
+
+### 1.1 - Organization or Department
+
+**Owning Organization**: [Full official name]
+
+### 1.2 - Team
+
+**Responsible Team**: [Team name and role]
+
+### 1.3 - Senior Responsible Owner
+
+**SRO Name**: [Name]
+**SRO Role**: [Job title]
+**SRO Accountability**: [Description of accountability]
+
+### 1.4 - External Supplier Involvement
+
+**External Suppliers Used**: [ ] Yes / [ ] No
+
+#### 1.4.1 - External Supplier(s)
+
+1. **Supplier Name**: [Company name]
+2. **Supplier Name**: [Company name]
+
+#### 1.4.2 - Companies House Number(s)
+
+1. [Companies House registration number]
+2. [Companies House registration number]
+
+#### 1.4.3 - External Supplier Role
+
+[Describe what each supplier provides - development, hosting, AI models, data processing, etc.]
+
+#### 1.4.4 - Procurement Procedure Type
+
+[ ] Open procedure
+[ ] Restricted procedure
+[ ] Competitive procedure with negotiation
+[ ] Competitive dialogue
+[ ] Innovation partnership
+[ ] Direct award (state justification): _______________
+[ ] Framework agreement: [Framework name, e.g., G-Cloud 14]
+
+#### 1.4.5 - Data Access Terms
+
+[Describe supplier access to data, data processing agreements, data residency, security controls]
+
+---
+
+## Section 2: Description and Rationale
+
+### 2.1 - Detailed Description
+
+**Technical Architecture**:
+[Describe the algorithmic tool's architecture, components, data flow]
+
+**Algorithm Type**:
+[ ] Rule-based (deterministic logic)
+[ ] Machine Learning - Supervised
+[ ] Machine Learning - Unsupervised
+[ ] Machine Learning - Reinforcement Learning
+[ ] Generative AI (LLM/Foundation Model)
+[ ] Computer Vision
+[ ] Natural Language Processing
+[ ] Optimization/Operations Research
+[ ] Statistical Model
+[ ] Hybrid approach
+[ ] Other: _______________
+
+**AI Model Details** (if applicable):
+
+- **Model Name/Type**: [e.g., GPT-4, BERT, custom model]
+- **Model Provider**: [e.g., OpenAI, Anthropic, Google, in-house]
+- **Model Version**: [Version number]
+- **Fine-tuned**: [ ] Yes / [ ] No
+- **Training Data**: [Description of training data used]
+
+**Technical Components**:
+
+- [List key technical components, APIs, databases, infrastructure]
+
+### 2.2 - Scope
+
+**Intended Use**:
+[Describe the specific use cases and boundaries]
+
+**Out of Scope**:
+[Explicitly state what the tool is NOT designed to do]
+
+**User Population**:
+
+- **Internal Users**: [Number and roles of staff using the tool]
+- **External Users**: [Number and types of citizens/public affected]
+
+**Geographic Scope**: [England/Scotland/Wales/NI/UK-wide/Specific regions]
+
+### 2.3 - Benefit
+
+**Intended Benefits**:
+[Describe expected positive outcomes]
+
+**Impact Metrics**:
+
+- [Metric 1: e.g., reduced processing time]
+- [Metric 2: e.g., improved accuracy]
+- [Metric 3: e.g., cost savings]
+
+**Evidence of Benefits**:
+[Cite evidence, trials, user research demonstrating benefits]
+
+### 2.4 - Previous Process
+
+**Before Implementation**:
+[Describe how this task was done before the algorithmic tool]
+
+**Comparison**:
+| Aspect | Previous Process | Current Tool |
+|--------|-----------------|--------------|
+| Time | | |
+| Accuracy | | |
+| Cost | | |
+| User Experience | | |
+| Fairness | | |
+
+### 2.5 - Alternatives Considered
+
+**Alternative 1**: [Option considered]
+
+- **Reason for Rejection**: [Why not chosen]
+
+**Alternative 2**: [Option considered]
+
+- **Reason for Rejection**: [Why not chosen]
+
+**Non-Algorithmic Approach**:
+[Why a non-AI/algorithmic solution was not suitable]
+
+---
+
+## Section 3: Decision-Making Process
+
+### 3.1 - Process Integration
+
+**Role in Workflow**:
+[Describe where the tool fits in the overall process]
+
+**Process Diagram**:
+
+```mermaid
+flowchart LR
+    A[Input] --> B[Processing Step 1]
+    B --> C[Algorithmic Tool]
+    C --> D[Processing Step 2]
+    D --> E[Output/Decision]
+```
+
+**Integration Points**:
+
+- [System 1 integration]
+- [System 2 integration]
+
+### 3.2 - Provided Information
+
+**Outputs**:
+[Describe what information the tool provides]
+
+**Output Format**:
+[ ] Score/Rating
+[ ] Classification/Category
+[ ] Recommendation
+[ ] Prediction
+[ ] Ranking/Prioritization
+[ ] Risk Assessment
+[ ] Generated Content
+[ ] Other: _______________
+
+**Output Interpretation**:
+[Explain how to interpret the tool's outputs]
+
+### 3.3 - Frequency and Scale of Usage
+
+**Usage Volume**:
+
+- **Decisions per day/month/year**: [Number]
+- **Users affected per day/month/year**: [Number]
+- **Staff users**: [Number]
+
+**Usage Pattern**:
+[ ] Continuous/Real-time
+[ ] Batch processing (daily/weekly/monthly)
+[ ] On-demand
+
+### 3.4 - Human Decisions and Review
+
+**Human Oversight Model**:
+[ ] Human-in-the-loop (review EVERY decision before action)
+[ ] Human-on-the-loop (periodic/sample review)
+[ ] Human-in-command (can override at any time)
+[ ] Fully automated (explain justification): _______________
+
+**Review Process**:
+[Describe how human reviewers assess algorithmic outputs]
+
+**Override Capability**:
+
+- **Can humans override**: [ ] Yes / [ ] No
+- **Override frequency**: [e.g., X% of cases]
+- **Override reasons**: [Common reasons for overriding algorithm]
+
+**Quality Assurance**:
+[Describe QA processes, sample checks, audits]
+
+### 3.5 - Required Training
+
+**Staff Training Program**:
+
+- **Duration**: [Hours/days of training]
+- **Content**: [What training covers]
+- **Certification**: [ ] Yes / [ ] No
+
+**Training Topics**:
+
+- [ ] How the algorithm works
+- [ ] AI limitations and risks
+- [ ] Bias and fairness awareness
+- [ ] Override procedures
+- [ ] Escalation process
+- [ ] User rights (contestability)
+
+**Ongoing Training**: [Frequency of refresher training]
+
+### 3.6 - Appeals and Contestability
+
+**Right to Contest**:
+[ ] Yes - users can contest algorithmic decisions
+[ ] No - not applicable (explain): _______________
+
+**Contest Process**:
+
+1. [Step 1: How users submit contest]
+2. [Step 2: Review process]
+3. [Step 3: Resolution]
+
+**Response Time**: [e.g., 28 days]
+
+**Human Review for Contested Decisions**:
+[Describe human review process for appeals]
+
+---
+
+## Section 4: Data
+
+### 4.1 - Data Sources
+
+**Input Data**:
+
+1. **Data Source 1**:
+   - **Type**: [e.g., Personal data, Administrative data, Open data]
+   - **Origin**: [Where data comes from]
+   - **Fields Used**: [Specific data fields]
+
+2. **Data Source 2**:
+   - **Type**: [e.g., Personal data, Administrative data, Open data]
+   - **Origin**: [Where data comes from]
+   - **Fields Used**: [Specific data fields]
+
+**Personal Data**: [ ] Yes / [ ] No
+
+**Special Category Data** (e.g., health, ethnicity, religion): [ ] Yes / [ ] No
+
+- If yes, specify: _______________
+- Legal basis: _______________
+
+### 4.2 - Data Sharing
+
+**Data Shared With**:
+
+- [Organization/Partner 1]: [Purpose]
+- [Organization/Partner 2]: [Purpose]
+
+**Legal Basis for Sharing**:
+[Cite specific legislation or data sharing agreements]
+
+### 4.3 - Data Quality and Maintenance
+
+**Data Quality Assurance**:
+[Describe data validation, cleansing, quality checks]
+
+**Data Freshness**:
+
+- **Update Frequency**: [Real-time / Daily / Weekly / Monthly]
+- **Historical Data**: [How far back data goes]
+
+**Data Completeness**:
+[Percentage complete, known gaps]
+
+### 4.4 - Data Storage and Security
+
+**Data Location**:
+[ ] UK
+[ ] EU
+[ ] USA
+[ ] Other: _______________
+
+**Cloud Provider**: [e.g., AWS, Azure, GCP, on-premise]
+
+**Security Measures**:
+
+- [ ] Encryption at rest
+- [ ] Encryption in transit
+- [ ] Access controls (RBAC)
+- [ ] Audit logging
+- [ ] Regular penetration testing
+- [ ] Cyber Essentials / Cyber Essentials Plus certified
+- [ ] ISO 27001 certified
+
+**Data Retention**:
+[How long data is kept and why]
+
+---
+
+## Section 5: Impact Assessments
+
+### 5.1 - Data Protection Impact Assessment (DPIA)
+
+**DPIA Completed**: [ ] Yes / [ ] No / [ ] In Progress
+
+**DPIA Date**: [YYYY-MM-DD]
+
+**DPIA Outcome**:
+[ ] Approved - low risk
+[ ] Approved with mitigations
+[ ] Requires ICO consultation
+
+**Key Risks Identified**:
+
+1. [Risk 1 and mitigation]
+2. [Risk 2 and mitigation]
+3. [Risk 3 and mitigation]
+
+**DPIA Review Date**: [YYYY-MM-DD]
+
+### 5.2 - Equality Impact Assessment (EqIA)
+
+**EqIA Completed**: [ ] Yes / [ ] No / [ ] In Progress
+
+**EqIA Date**: [YYYY-MM-DD]
+
+**Protected Characteristics Assessed**:
+
+- [ ] Age
+- [ ] Disability
+- [ ] Gender reassignment
+- [ ] Marriage and civil partnership
+- [ ] Pregnancy and maternity
+- [ ] Race
+- [ ] Religion or belief
+- [ ] Sex
+- [ ] Sexual orientation
+
+**Impact Assessment**:
+| Characteristic | Potential Impact | Mitigation |
+|----------------|------------------|------------|
+| Age | | |
+| Disability | | |
+| Race | | |
+| Sex | | |
+
+**EqIA Outcome**: [Summary and actions]
+
+**EqIA Review Date**: [YYYY-MM-DD]
+
+### 5.3 - Human Rights Assessment
+
+**Human Rights Assessment Completed**: [ ] Yes / [ ] No
+
+**ECHR Articles Considered**:
+
+- [ ] Article 6: Right to fair trial
+- [ ] Article 8: Right to privacy
+- [ ] Article 14: Freedom from discrimination
+- [ ] Other: _______________
+
+**Human Rights Safeguards**:
+[Describe how human rights are protected]
+
+### 5.4 - Other Impact Assessments
+
+**Environmental Impact Assessment**: [ ] Yes / [ ] No
+
+- **Carbon Footprint**: [e.g., CO2e from model training/inference]
+- **Sustainability Measures**: [Green hosting, energy efficiency]
+
+**Accessibility Assessment** (WCAG 2.2 Level AA): [ ] Yes / [ ] No
+
+**Security Risk Assessment**: [ ] Yes / [ ] No
+
+---
+
+## Section 6: Fairness, Bias, and Discrimination
+
+### 6.1 - Bias Assessment
+
+**Bias Testing Completed**: [ ] Yes / [ ] No / [ ] In Progress
+
+**Bias Testing Date**: [YYYY-MM-DD]
+
+**Testing Methodology**:
+[Describe how bias was tested - datasets, metrics, methods]
+
+### 6.2 - Fairness Metrics
+
+**Fairness Metrics Calculated**:
+
+- [ ] Demographic parity
+- [ ] Equalized odds
+- [ ] Equal opportunity
+- [ ] Calibration
+- [ ] Other: _______________
+
+**Results**:
+| Protected Characteristic | Metric | Result | Threshold | Pass/Fail |
+|--------------------------|--------|--------|-----------|-----------|
+| Gender | | | | |
+| Ethnicity | | | | |
+| Age | | | | |
+| Disability | | | | |
+
+### 6.3 - Known Limitations and Bias
+
+**Known Biases**:
+
+1. [Bias 1]: [Description and cause]
+2. [Bias 2]: [Description and cause]
+3. [Bias 3]: [Description and cause]
+
+**Mitigation Strategies**:
+
+- [Strategy 1]
+- [Strategy 2]
+- [Strategy 3]
+
+**Residual Risk**:
+[Describe any remaining bias that couldn't be fully eliminated]
+
+### 6.4 - Training Data Bias
+
+**Training Data Review**: [ ] Yes / [ ] No
+
+**Training Data Demographics**:
+[Describe representation of different groups in training data]
+
+**Known Data Gaps**:
+[Groups underrepresented or missing from training data]
+
+**Data Augmentation**:
+[Methods used to address data gaps]
+
+### 6.5 - Ongoing Bias Monitoring
+
+**Monitoring Frequency**: [Daily / Weekly / Monthly / Quarterly / Annually]
+
+**Monitoring Metrics**:
+
+- [Metric 1]
+- [Metric 2]
+
+**Alert Thresholds**:
+[When bias monitoring triggers a review]
+
+---
+
+## Section 7: Technical Details
+
+### 7.1 - Model Performance
+
+**Performance Metrics**:
+| Metric | Value | Benchmark |
+|--------|-------|-----------|
+| Accuracy | | |
+| Precision | | |
+| Recall | | |
+| F1 Score | | |
+| False Positive Rate | | |
+| False Negative Rate | | |
+
+**Performance by Demographic Group**:
+[Break down metrics by protected characteristics]
+
+### 7.2 - Model Explainability
+
+**Explainability Approach**:
+[ ] SHAP (SHapley Additive exPlanations)
+[ ] LIME (Local Interpretable Model-agnostic Explanations)
+[ ] Feature importance
+[ ] Decision tree visualization
+[ ] Natural language explanations
+[ ] Other: _______________
+[ ] Not applicable (rule-based system)
+
+**Explanation Provided to Users**: [ ] Yes / [ ] No
+
+**Example Explanation**:
+[Provide example of how decision is explained to user]
+
+### 7.3 - Model Versioning and Change Management
+
+**Current Model Version**: [Version number]
+
+**Version History**:
+| Version | Date | Changes | Impact |
+|---------|------|---------|--------|
+| 1.0 | | Initial deployment | |
+| 1.1 | | | |
+| 2.0 | | | |
+
+**Change Management Process**:
+[Describe how model updates are tested, approved, deployed]
+
+**Rollback Capability**: [ ] Yes / [ ] No
+
+### 7.4 - Model Monitoring and Drift Detection
+
+**Drift Monitoring**: [ ] Yes / [ ] No
+
+**Monitoring Metrics**:
+
+- **Data Drift**: [How input data distribution is monitored]
+- **Concept Drift**: [How relationship between inputs and outputs is monitored]
+- **Performance Drift**: [How accuracy changes over time]
+
+**Alert Thresholds**:
+[When drift triggers retraining or review]
+
+**Retraining Schedule**:
+[ ] Triggered by drift detection
+[ ] Fixed schedule: [Frequency]
+[ ] Ad-hoc based on performance review
+
+---
+
+## Section 8: Testing and Assurance
+
+### 8.1 - Testing Approach
+
+**Testing Phases**:
+
+- [ ] Unit testing (individual components)
+- [ ] Integration testing (system interaction)
+- [ ] User acceptance testing (UAT)
+- [ ] A/B testing
+- [ ] Shadow testing (parallel running)
+- [ ] Red teaming (adversarial testing)
+
+**Test Coverage**: [Percentage or description]
+
+### 8.2 - Edge Cases and Failure Modes
+
+**Known Edge Cases**:
+
+1. [Edge case 1 and handling]
+2. [Edge case 2 and handling]
+3. [Edge case 3 and handling]
+
+**Failure Modes**:
+| Failure Mode | Probability | Impact | Mitigation |
+|--------------|-------------|--------|------------|
+| Model returns no result | | | |
+| Model fails to process input | | | |
+| Performance degradation | | | |
+
+**Fallback Procedures**:
+[What happens when algorithm fails]
+
+### 8.3 - Security Testing
+
+**Security Assessments Completed**:
+
+- [ ] Penetration testing
+- [ ] Vulnerability scanning
+- [ ] AI-specific threat assessment:
+  - [ ] Prompt injection testing (for LLMs)
+  - [ ] Data poisoning risk assessment
+  - [ ] Model inversion attack assessment
+  - [ ] Adversarial example testing
+  - [ ] Model theft/extraction risk
+
+**Security Findings**: [Summary and remediation]
+
+### 8.4 - Independent Assurance
+
+**Independent Review**: [ ] Yes / [ ] No
+
+**Reviewer**: [Organization/team]
+
+**Review Date**: [YYYY-MM-DD]
+
+**Review Outcome**: [Summary and recommendations]
+
+**External Audit**: [ ] Yes / [ ] No / [ ] Planned
+
+---
+
+## Section 9: Transparency and Explainability
+
+### 9.1 - Public Information
+
+**Public Disclosure**:
+
+- [ ] Tool is publicly disclosed
+- [ ] ATRS record published on GOV.UK
+- [ ] Information on department website
+- [ ] Model card published
+- [ ] Open source code
+
+**Website URL**: [URL to public information]
+
+### 9.2 - User Communication
+
+**Users Informed**: [ ] Yes / [ ] Partially / [ ] No
+
+**How Users Are Informed**:
+
+- [ ] Direct notification
+- [ ] Website information
+- [ ] Privacy notice
+- [ ] Terms of service
+- [ ] In-application message
+
+**Information Provided to Users**:
+
+- [ ] That an algorithm is being used
+- [ ] What the algorithm does
+- [ ] How it affects them
+- [ ] How to contest decisions
+- [ ] How their data is used
+
+### 9.3 - Model Card
+
+**Model Card Published**: [ ] Yes / [ ] No
+
+**Model Card URL**: [URL]
+
+**Model Card Contents**:
+
+- [ ] Model architecture
+- [ ] Training data description
+- [ ] Performance metrics
+- [ ] Limitations
+- [ ] Bias testing results
+- [ ] Intended use
+- [ ] Out-of-scope uses
+
+---
+
+## Section 10: Governance and Oversight
+
+### 10.1 - Governance Structure
+
+**Governance Board**: [Name of board/committee]
+
+**Board Composition**:
+
+- [Role 1]
+- [Role 2]
+- [Role 3]
+
+**Board Responsibilities**:
+
+- [Responsibility 1]
+- [Responsibility 2]
+
+**Meeting Frequency**: [e.g., Monthly, Quarterly]
+
+### 10.2 - Risk Management
+
+**Risk Register Maintained**: [ ] Yes / [ ] No
+
+**Top Risks**:
+| Risk | Likelihood | Impact | Mitigation | Owner |
+|------|------------|--------|------------|-------|
+| | | | | |
+| | | | | |
+| | | | | |
+
+**Risk Review Frequency**: [e.g., Monthly]
+
+### 10.3 - Incident Management
+
+**Incident Response Plan**: [ ] Yes / [ ] No
+
+**Incident Types**:
+
+- [ ] Algorithm error/failure
+- [ ] Bias/discrimination incident
+- [ ] Data breach
+- [ ] Security incident
+- [ ] Public complaint
+
+**Incident Response Process**:
+
+1. [Detection]
+2. [Escalation]
+3. [Response]
+4. [Resolution]
+5. [Review and lessons learned]
+
+**Contact for Incidents**: [email@department.gov.uk]
+
+### 10.4 - Audit Trail
+
+**Audit Logging**: [ ] Yes / [ ] No
+
+**Logged Events**:
+
+- [ ] All algorithmic decisions
+- [ ] User access
+- [ ] Configuration changes
+- [ ] Model updates
+- [ ] Data access
+
+**Log Retention**: [Duration]
+
+**Log Review**: [Frequency and process]
+
+---
+
+## Section 11: Compliance
+
+### 11.1 - Legal Basis
+
+**Primary Legislation**:
+
+- [Act 1]: [Specific sections]
+- [Act 2]: [Specific sections]
+
+**Regulatory Compliance**:
+
+- [ ] UK GDPR
+- [ ] Data Protection Act 2018
+- [ ] Equality Act 2010
+- [ ] Human Rights Act 1998
+- [ ] Freedom of Information Act 2000
+- [ ] Public Sector Equality Duty
+- [ ] AI-specific regulation (if applicable)
+
+### 11.2 - Data Protection
+
+**Data Controller**: [Organization name]
+
+**Data Protection Officer**: [Name and contact]
+
+**ICO Registration Number**: [Number]
+
+**Legal Basis for Processing**:
+[ ] Consent
+[ ] Contract
+[ ] Legal obligation
+[ ] Vital interests
+[ ] Public task
+[ ] Legitimate interests
+
+### 11.3 - Standards Compliance
+
+**Standards Followed**:
+
+- [ ] Technology Code of Practice
+- [ ] Government Design Principles
+- [ ] GDS Service Standard
+- [ ] Data Ethics Framework
+- [ ] AI Governance Standards
+- [ ] ISO 27001 (Information Security)
+- [ ] ISO 9001 (Quality Management)
+- [ ] Other: _______________
+
+### 11.4 - Procurement Compliance
+
+**Procurement Route**: [e.g., G-Cloud, DOS, Open Tender]
+
+**Contract Value**: [£ amount or range]
+
+**IR35 Compliance** (if contractors): [ ] Yes / [ ] No / [ ] N/A
+
+---
+
+## Section 12: Performance and Outcomes
+
+### 12.1 - Success Metrics
+
+**KPIs**:
+
+1. [KPI 1]: [Target] → [Current performance]
+2. [KPI 2]: [Target] → [Current performance]
+3. [KPI 3]: [Target] → [Current performance]
+
+**Benefits Realised**:
+
+- [Benefit 1]: [Evidence]
+- [Benefit 2]: [Evidence]
+- [Benefit 3]: [Evidence]
+
+### 12.2 - User Feedback
+
+**User Research Conducted**: [ ] Yes / [ ] No
+
+**User Satisfaction**: [Score or qualitative feedback]
+
+**User Concerns**:
+
+- [Concern 1 and response]
+- [Concern 2 and response]
+
+**Feedback Mechanism**: [How users can provide feedback]
+
+### 12.3 - Continuous Improvement
+
+**Improvement Log**:
+| Date | Issue | Action Taken | Outcome |
+|------|-------|--------------|---------|
+| | | | |
+| | | | |
+
+**Lessons Learned**:
+
+- [Lesson 1]
+- [Lesson 2]
+- [Lesson 3]
+
+---
+
+## Section 13: Review and Updates
+
+### 13.1 - Review Schedule
+
+**ATRS Review Frequency**: [e.g., Annually, or when significant changes]
+
+**Next Review Date**: [YYYY-MM-DD]
+
+**Trigger for Unscheduled Review**:
+
+- [ ] Significant model change
+- [ ] Change in usage/scope
+- [ ] Incident or complaint
+- [ ] Regulatory change
+- [ ] New evidence of bias
+
+### 13.2 - Version History
+
+| Version | Date | Author | Changes |
+|---------|------|--------|---------|
+| 1.0 | | | Initial ATRS record |
+| 1.1 | | | |
+| 2.0 | | | |
+
+### 13.3 - Contact for Updates
+
+**ATRS Record Owner**: [Name]
+**Email**: [email@department.gov.uk]
+**Last Updated**: [YYYY-MM-DD]
+
+---
+
+## Appendices
+
+### Appendix A: Glossary
+
+- **Algorithm**: [Definition for your context]
+- **Machine Learning**: [Definition]
+- **Training Data**: [Definition]
+- **Bias**: [Definition]
+- **DPIA**: Data Protection Impact Assessment
+- **EqIA**: Equality Impact Assessment
+
+### Appendix B: References
+
+- [Reference 1]
+- [Reference 2]
+- [Reference 3]
+
+### Appendix C: Supporting Documentation
+
+- [ ] Full DPIA
+- [ ] Full EqIA
+- [ ] Human Rights Assessment
+- [ ] Model Card
+- [ ] User Research Reports
+- [ ] Audit Reports
+- [ ] Test Results
+
+---
+
+## Publication Checklist
+
+Before publishing this ATRS record:
+
+**Completeness**:
+
+- [ ] All Tier 1 fields completed
+- [ ] All mandatory Tier 2 fields completed
+- [ ] Plain English used in Tier 1
+- [ ] Technical detail sufficient in Tier 2
+
+**Approvals**:
+
+- [ ] Senior Responsible Owner approval
+- [ ] Legal/compliance review
+- [ ] Data Protection Officer review
+- [ ] Communications team review (public-facing content)
+
+**Impact Assessments**:
+
+- [ ] DPIA completed and approved
+- [ ] EqIA completed
+- [ ] Human Rights Assessment completed (if applicable)
+- [ ] Security Risk Assessment completed
+
+**Quality**:
+
+- [ ] Accuracy verified
+- [ ] Links work
+- [ ] Contact details current
+- [ ] Sensitive information redacted
+
+**Publication**:
+
+- [ ] Published on GOV.UK ATRS repository
+- [ ] Published on department website
+- [ ] Stakeholders informed
+- [ ] Review date set
+
+---
+
+**END OF ATRS RECORD**
+
+For guidance on completing this template, see:
+https://www.gov.uk/government/publications/guidance-for-organisations-using-the-algorithmic-transparency-recording-standard
+
+For questions: algorithmic-transparency@dsit.gov.uk
+
+## External References
+
+> This section provides traceability from generated content back to source documents.
+> Follow citation instructions in the project's citation reference guide.
+
+### Document Register
+
+| Doc ID | Filename | Type | Source Location | Description |
+|--------|----------|------|-----------------|-------------|
+| *None provided* | — | — | — | — |
+
+### Citations
+
+| Citation ID | Doc ID | Page/Section | Category | Quoted Passage |
+|-------------|--------|--------------|----------|----------------|
+| — | — | — | — | — |
+
+### Unreferenced Documents
+
+| Filename | Source Location | Reason |
+|----------|-----------------|--------|
+| — | — | — |
+
+---
+
+**Generated by**: ArcKit `/arckit:atrs` command
+**Generated on**: [DATE]
+**ArcKit Version**: [VERSION]
+**Project**: [PROJECT_NAME]
+**Model**: [AI_MODEL]