<< All versions

Skill v1.0.0

currentAutomated scan100/100
tiger-ai-lab/rewardharness/enforce-logical-consistency
──Details
PublishedNovember 3, 2026 at 09:35 AM
Content Hashsha256:bfcf156df6437dae...
Git SHA0c1e7051ab4f
──Files
Files (1 file, 1.1 KB)
SKILL.md1.1 KBactive
SKILL.md · 20 lines · 1.1 KB

schema_version: 2 name: enforce-logical-consistency kind: skill description: Ensures the reasoning chain does not contain internal contradictions between stated facts and image evaluations. metadata: version: "1.0.0"


Enforce Logical Consistency

When evaluating images, your reasoning must remain logically consistent from start to finish.

  1. Establish Facts Clearly: If you define a factual premise or aesthetic (e.g., "New York subways are typically industrial, dirty, and utilitarian"), you must anchor your evaluation to this premise.
  2. Avoid Direct Contradictions: Do not state a fact and then reward an image for the exact opposite. For example, you cannot praise an image for being "sleek, modern, and metallic" and then claim this makes it "closer to the typical New York subway appearance."
  3. Check Your Conclusion: Before finalizing your preference, review your reasoning. Does your final choice actually embody the characteristics you defined as correct at the beginning of your analysis? If not, adjust your preference or re-evaluate the images.

<!-- Release review: RewardHarness v0.4.1 (2026-08-26). -->

All versions