Selecting Enforcements

Enforcement Selection

Guide users through selecting enforcement patterns that ensure agent system lives its values

Enforcement Selection Skill

Overview

This skill guides users through selecting enforcement patterns that will ensure their agent system lives its values, not just states them.

Philosophy

Values without enforcement are just aspirations.

Enforcement patterns translate our aspirations into operational reality. They are the mechanisms that ensure our values are practiced, not just professed.

Purpose

Help users select and configure enforcement patterns that match their project's needs and risk profile, ensuring the agent system maintains integrity through measurable, enforceable standards.

Trigger

• During onboarding after principles are defined
• User mentions "enforcement", "quality gates", "standards"
• Request to configure automated checks

Process

1. Review the task requirements.
2. Apply the skill's methodology.
3. Validate the output against the defined criteria.

Step 1: Explain Enforcement Philosophy

Your axioms and principles define what you value.
Enforcement patterns ensure those values are lived.

Think of enforcement as:
- Quality: Craftsmanship standards that code must meet
- Safety: Guardrails that protect users and systems
- Integrity: Checks that keep you true to your purpose

Excellence emerges from consistent discipline.

Step 2: Present Enforcement Categories

Select enforcement patterns by category:

QUALITY ENFORCEMENT (Craftsmanship)
[x] E1: Test Coverage Gate - Verify code is tested
[x] E2: Peer Review Gate - Ensure collaborative refinement
[ ] E3: Documentation Completeness - Share knowledge clearly
[x] E4: Style Consistency - Harmony in the details

SAFETY ENFORCEMENT (Protection)
[x] E5: Destructive Action Confirmation - Prevent accidental harm
[ ] E6: Backup Before Modification - Enable recovery
[ ] E7: Security Vulnerability Check - Protect against threats
[ ] E8: Production Safeguard - Extra care for live systems

INTEGRITY ENFORCEMENT (Alignment)
[x] E9: Axiom Compliance Check - Rules trace to axioms
[ ] E10: Purpose Alignment Check - Work serves mission
[ ] E11: Transparency Log - Decisions are documented

Recommended minimum: E1, E2, E4, E5, E9

Step 3: Configure Thresholds

For each selected enforcement, configure thresholds:

E1 - Test Coverage Gate:
  Minimum coverage: [80]%
  New code coverage: [90]%
  Severity: [blocking]

E2 - Peer Review Gate:
  Minimum reviewers: [1]
  Approval required: [yes]
  Comments must be resolved: [yes]

(continue for each selected enforcement...)

Step 4: Configure Override Policies

Override policies define when rules can be bypassed:

E1 (Test Coverage):
  [ ] Never - Must always pass
  [x] With justification - Document why
  [ ] With approval - Requires senior sign-off

E5 (Destructive Confirmation):
  [x] Never - Always require confirmation
  [ ] With justification
  [ ] With approval

For safety-critical projects, recommend "Never" or "With approval" for most enforcements.

Step 5: Generate Enforcement Configuration

# enforcement.yaml - Generated enforcement configuration

enforcements:
  quality:
    E1_test_coverage:
      enabled: true
      threshold: 80
      severity: blocking
      override: with_justification

    E2_code_review:
      enabled: true
      min_reviewers: 1
      require_approval: true
      override: with_approval

    E4_style_consistency:
      enabled: true
      severity: blocking
      override: never

  safety:
    E5_destructive_confirmation:
      enabled: true
      severity: blocking
      override: never

    E8_production_safeguard:
      enabled: true
      require_staging_test: true
      override: with_approval

  integrity:
    E9_axiom_compliance:
      enabled: true
      severity: blocking
      override: never

Outputs

1. Enforcement Configuration - enforcement.yaml for generated project
2. Integration with CI/CD - Hooks for automated enforcement
3. Override Log Template - For documenting exceptions

Integration Points

• Input from: Axiom selection, principle definition
• Reads: {directories.patterns}/enforcement/*.json
• Outputs to: enforcement.yaml, CI/CD configuration
• Related: practice-selection skill

Best Practices

1. Start with essential enforcements (E1, E2, E5, E9)
2. Match enforcement strictness to project risk level
3. "Never" override for safety-critical enforcements
4. Document rationale for all override policies
5. Review enforcement effectiveness quarterly

When to Use

This skill should be used when strict adherence to the defined process is required.

Prerequisites

• Basic understanding of the agent factory context.
• Access to the necessary tools and resources.