| id | DEV-440 | |
|---|---|---|
| title | Judge Work by Value Delivered, Not Effort Spent | |
| status | active | |
| enforcement | manual | |
| severity | error | |
| depends_on |
|
Work gets justified by the hours it took, so mechanical tasks a tool does in minutes still bill full cost, and effort stands in for value.
Measure a change by the value it delivers against what it cost to produce, not by the time it took. High effort on low-value work is waste, not merit.
- Before spending hours on a task, ask whether the output justifies the cost. A day designing a valuable solution is fine; hours porting a file between formats or fighting a linter are not.
- Use AI to do the mechanical work: format conversions, boilerplate, refactors, lint fixes. Reserve human hours for judgment and design.
- When effort is high relative to value, say so and cut it, rather than defending the hours after the fact.
Good: a spec is migrated between formats with AI in minutes, and the reported time reflects that. Bad: the same migration is done by hand over hours and billed as if the effort were the value.
- Work is assessed by value against cost, not by hours spent
- Mechanical, tool-automatable work is not billed as high-effort human work
- High-effort, low-value work is flagged and cut rather than justified