Dali by Lulu

Several tools cluster around the same jobs: community_benchmark, creative_patterns, enhancement_path, and prompt_neighbors all surface winning or improvement patterns, while score_creative and score_creative_from_view are near-twins differentiated mainly by input method. The detailed descriptions usually resolve the ambiguity, but an agent skimming names alone could easily misselect.

Roughly two-thirds of the tools follow a clear verb_noun convention (analyze_*, score_*, list_*, suggest_*, track_*, enhance_*), but the remaining tools are bare noun phrases (community_benchmark, creative_patterns, dali_version, enhancement_path, my_story, prompt_neighbors). The snake_case style is consistent, but the action-oriented vs. noun-oriented split makes the pattern mixed.

At 16 tools, the surface sits at the upper edge of the reasonable range and feels slightly heavy for the core prompt/image scoring mission. Some consolidation is possible among the community-pattern tools and the two image-scoring variants, though each tool has a defensible role.

The core lifecycle is well covered: discover generators, score prompts or images, get enhancement briefs, track improvements, and consult community/personal analytics. Minor gaps exist—such as no single-prompt cross-generator comparison and no management of tracked enhancement history—but they do not create dead ends.