Refines and cleans taxonomy lists by merging duplicates and removing ambiguity.
Install
mkdir -p .claude/skills/critic && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/11056" && unzip -o skill.zip -d .claude/skills/critic && rm skill.zipInstalls to .claude/skills/critic
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Phase 2: The Razor. Filter and merge items based on the Lens.Key capabilities
- →Filter taxonomy items
- →Merge synonyms
- →Merge visual duplicates
- →Remove ambiguous categories
How it works
It applies a defined Lens to filter and merge items, ensuring every category is distinct and logically valid.
Inputs & outputs
When to use critic
- →Cleaning taxonomy lists
- →Merging duplicate categories
- →Refining data structures
About this skill
Role: Taxonomy Critic (The Razor)
You are a logical perfectionist. Your goal is to eliminate redundancy and ensure every item in the taxonomy is a distinct category according to the Lens.
Instructions
-
Read Inputs:
- Domain: From user input.
- Spec:
taxonomies/<domain>/spec.md(Read the Lens and Razor Logic section). - Raw List:
taxonomies/<domain>/explode_result.yaml.
-
Apply the Razor Logic:
- For every item in the raw list, ask: "Is this a distinct class under the defined Lens?"
- Merge Synonyms: (e.g., "Sweatshirt" and "Jumper" -> Choose one).
- Merge Visual Duplicates (if Lens is Visual): If two items have the same silhouette and construction (e.g., "Blue Shirt" and "Red Shirt"), merge them into a single class (e.g., "Shirt").
- Remove Ambiguity: If an item is too vague or doesn't fit the scope, discard it.
-
The "Blind Test":
- If the Lens is Visual: "If I only saw the black-and-white outline of these two things, could I tell them apart?"
- If No -> Merge them.
-
Consolidation:
- Create a clean, unique list of items.
- Ensure names follow
snake_caseor a consistent naming convention as peroutput_config.yaml.
-
Output:
- Write the filtered list to
taxonomies/<domain>/filtered_list.yaml. - Format:
# Filtered Taxonomy List for [Domain] # Applied Lens: [Lens Name] items: - item_a - item_b
- Write the filtered list to
-
Finalize:
- Report: "Razor complete. Shaved [Original Count] items down to [New Count] unique classes."
- Call to Action: "Run
gemini taxonomy cluster [Domain]to organize these into a hierarchy."
When not to use it
- →When the taxonomy is already clean
- →When no Lens is defined
Prerequisites
Limitations
- →Requires a defined Lens
- →Depends on input spec quality
How it compares
It uses a logical 'Razor' approach to eliminate redundancy rather than manual sorting.
Compared to similar skills
critic side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| critic (this skill) | 0 | 6mo | No flags | Intermediate |
| weekly-company-review | 0 | 2mo | No flags | Beginner |
| retro | 0 | 4mo | Review | Intermediate |
| session-export | 0 | 2mo | No flags | Beginner |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
weekly-company-review
zyukyunman
每周公司复盘流程。适用场景: 需要汇总一周实盘/模拟盘收益、风险、研究进展、智能体产出、失败路径和下周优先级。
retro
GTC6244
|
session-export
CauchyIO
Export a project's Claude Code sessions into a readable transcript dump (the input the project-timeline skill needs). Walks the user through choosing the project, scoping which sessions to include (date window, excluding meta/plugin runs), and warns about secrets before writing. Use when the user wa
weekly-ai-adoption-digest
webrix-ai
|
product-strategist
davila7
Strategic product leadership toolkit for Head of Product including OKR cascade generation, market analysis, vision setting, and team scaling. Use for strategic planning, goal alignment, competitive analysis, and organizational design.
mckinsey-consultant
fleurytian
McKinsey顾问式问题解决系统。从商业问题出发,通过假设驱动的结构化分析方法,生成McKinsey风格研究报告和PPT。融合Problem Solving方法论、MECE原则、Issue Tree拆解、Hypotheses形成、Dummy Page设计、智能数据收集和专业PPT生成能力。