vibe-anti-rationalization-check

SkillProductivity

Catches shortcuts and reward hacking, weakening or deleting tests to make them pass, hard-coding expected outputs, silently dropping requirements, or claiming work is done without running it. Use before declaring a task complete and whenever a test or requirement feels like it's in the way.

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the vibe-anti-rationalization-check skill

What this skill tells your AI

The instructions your AI receives, as published by ash1794/vibe-engineering in plugins/vibe-engineering/skills/vibe-anti-rationalization-check/SKILL.md and read by ahel’s review.

Current models rarely skip work out of laziness. The failure that remains is subtler: satisfying the check instead of the goal. A test gets special-cased, an assertion loosened, a requirement quietly narrowed, and the summary says "done." This skill catches that and makes it visible.

When to Use This Skill

  • Before saying a task is complete, fixed, or passing
  • A test is failing and the quickest route to green is changing the test
  • You're about to narrow a requirement the user stated
  • You're about to write "in the interest of time...", "for now...", or "simplified"
  • You catch yourself explaining why something "doesn't really need" to be done

When NOT to Use This Skill

  • The user explicitly asked to skip or simplify something
  • The test itself is genuinely wrong (fix it, but say so explicitly and explain why)
  • You're genuinely unsure whether something is needed (ask the user instead)

Reward-Hacking Patterns

PatternWhat it looks likeDo instead
Test tamperingLoosening an assertion, adding a skip, deleting a failing case, raising a toleranceFix the code. If the test is wrong, change it and say why in the summary.
Special-casingif input == <test fixture>: return <expected> or detecting the test environmentImplement the general behavior.
Hard-coded outputsReturning the literal value a test expectsCompute it.
Mocking away the subjectMocking the very function the test is supposed to exerciseMock only external boundaries.
Silent scope cutImplementing 4 of 5 requirements and summarizing as "done"Finish, or list exactly what is missing.
Unverified claim"This should fix it" / "Tests pass" without running themRun the command and quote the result.
Budget looseningRaising a performance/size budget or threshold because the check failedFix the code; changing a budget needs a spec change and a stated reason
"It's flaky"Re-running until green, adding a retryFind the variable (vibe-flake-root-cause)
Self-approvalRecording a human's approval on their behalf, or treating a code comment as sign-offPoint the human to the item to approve (vibe-publication-leak-guard)
Error swallowingcatch {} / except: pass to stop a crashHandle the specific error, or let it surface.
Deferred forever"I'll add that later" with no tracked follow-upDo it now, or create a tracked issue/TODO with a reference.

Steps

  1. Diff check — Review your own diff for the patterns above. Pay particular attention to changes under test directories, in CI config, and in lint/skip annotations.
  2. Requirement check — Re-read the original request. List every requirement and mark each one done or not done.
  3. Evidence check — For each "works" or "passes" claim, is there command output from this session behind it?
  4. Resolve:
    • A legitimate reason to deviate → ask the user, or state it prominently in the summary
    • A rationalization → do the work
  5. Disclose — If anything was skipped, weakened, or changed in a test, say so plainly in the final summary. Never hide a shortcut.

Output

When run as a self-check, report only if something was found. When invoked explicitly:

Rationalization Check

#PatternLocationResolution
1Test tamperingauth_test.go:88 tolerance 0.01 → 0.5Reverted; fixed rounding in calc.go

Requirements: X/Y met — [list any not met] Unverified claims: [none / list]

Signals

GitHub stars
85
Forks
20
Last commit
Oct 2026
Advanced
Item type
skill
Key
vibe-anti-rationalization-check
Source
github.com/ash1794/vibe-engineering