Run the relevant verification command before claiming a result.

Workflow

•Identify what command proves the claim
•Run the command (fresh, not a previous run)
•Read the full output — check exit code, count failures
•Report actual results with evidence

Claim-to-Evidence Map

Claim	Required Evidence
Tests pass	Test command output showing 0 failures
Linter clean	`golangci-lint run` output showing 0 errors
Build succeeds	`go build` exit 0 (linter passing is not enough)
Bug fixed	Original symptom no longer reproduces
Regression tested	Red-green cycle: test fails without fix, passes with it
All checks pass	`make audit` output showing all steps pass

Transform Vague Tasks into Verifiable Goals

Before starting, rewrite the task as a testable outcome:

Task as given	Verifiable goal
"Add validation"	Write tests for invalid inputs, then make them pass
"Fix the bug"	Write a test that reproduces it, then make it pass
"Refactor X"	Ensure tests pass before and after
"Improve performance"	Measure before, change, measure after, compare

For multi-step work, pair each step with its check:

code

1. [Step] → verify: [check]
2. [Step] → verify: [check]

Strong success criteria let you loop independently. Weak criteria ("make it work") require constant clarification.

Examples

code

Good:  ran `make audit` → "All checks pass (format, vet, lint, test)"
Good:  ran `go test ./...` → "34/34 tests pass"
Avoid: "Should pass now" / "Looks correct" (without running anything)

Relationship to /qa

/qa tells you what to run. /verify reminds you to actually run it before claiming the result.