<< All versions

Skill v1.0.0

currentLLM-judged scan90/100
tiger-ai-lab/rewardharness/artist-and-style-knowledge
──Details
PublishedOctober 16, 2026 at 10:29 AM
Content Hashsha256:fe5b00a7959d2f45...
Git SHA0c1e7051ab4f
──Files
Files (1 file, 1.2 KB)
SKILL.md1.2 KBactive
SKILL.md · 19 lines · 1.2 KB

version: "1.0.0" schema_version: 2 name: artist-and-style-knowledge kind: skill description: Guidelines for evaluating edits that reference specific artists, art movements, or cultural references.


When an instruction references a specific artist (e.g., Claes Oldenburg, Gustav Klimt, Michelangelo) or style:

  1. Identify the Core Characteristics: Recall the defining features of the artist's work.
  • Claes Oldenburg: Oversized replicas of everyday objects (e.g., giant burgers, spoons) placed in environments.
  • Gustav Klimt: Gold leaf, intricate geometric patterns, romantic/symbolist portraiture.
  • Michelangelo: High Renaissance, muscular and idealized human forms, fresco textures, classical composition.
  1. Evaluate Semantic Accuracy: Does the edited image incorporate these specific defining features, or does it just apply a generic filter? (e.g., an Oldenburg edit should show a giant everyday object, not an abstract sculpture).
  2. Avoid Generic Interpretations: Penalize images that fail to capture the specific artist's signature style and instead use a generic 'painting' or 'sculpture' effect.

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

All versions