Verifiable Tasks Principle

Rule

Prioritize complex tasks with strong verification mechanisms over simpler tasks without objective feedback.

Context

When selecting autonomous work, especially for complex or high-value tasks.

Detection

Observable signals indicating poor task selection:

• Avoiding complex tasks because "too hard for autonomous"
• Selecting simple tasks just because they seem "safe"
• Working on subjective tasks without clear completion criteria
• Spending time on work where success/failure is ambiguous

Pattern

Choose based on verifiability, not complexity:

HIGH Verifiability (Choose):
  - Implement feature with test suite
  - Fix bug with failing tests
  - Refactor with type checking + tests
  - Performance optimization with benchmarks

LOW Verifiability (Deprioritize):
  - Creative writing (subjective)
  - Design decisions (judgment)
  - Strategic planning (needs input)

Key insight: Tests + CI + Type checking = Objective verification → Autonomy-friendly

Outcome

Following this principle enables:

• Autonomous work on complex problems (with verification)
• Quality output through iterative improvement
• Confidence in results (tests pass = done)
• Higher value delivery (complex > simple)

Related

• Code Coverage Principle - Tests enable complex autonomy