Garry Tan's Review Skill
Description
Recommended by
nene@YouMind
Why we love this skill
Centered on Garry Tan’s rigorous entrepreneurial and builder modes, this Skill provides CEO-level product review through six sharp questions and five generative questions, helping you validate ideas and optimize plans.
Based on gstack's office-hours and plan-ceo-review, this Skill is adapted for the YouMind environment for product review and document generation.
Related Skills
View all
SKILLS Refinement Workshop
Seven processes, one closed loop. SkillForge Refinement Workshop transforms the gstack engineering sprint methodology open-sourced by YC leader Garry Tan into a full lifecycle management system for Claude Code Skills. From six soul-searching questions to verify demand authenticity, to architecture planning, SKILL.md forging, quality review, eval testing, release packaging, and finally review and iteration—each stage has dedicated AI roles, with confidence scores and anti-spam mechanisms. It also includes two additional processes: emergency diagnosis and security audit. No 'good ideas,' only evidence-based judgments. Every skill is tempered through seven processes.
Skill Designer
Can't write a Skill? Just describe it in a sentence. Skill Designer is a 'Skill that teaches you to write Skills' — it uses a 7-layer meta-framework (Intent Definition → Input Classification → Workflow Design → Tool Strategy → Output Specification → Constraints and Boundaries → Composability & Testing) to turn your vague idea into a well-structured, immediately executable YouMind Skill. It supports two paths: Design from Scratch (describe your idea, it automatically asks for missing info, runs through the 7 layers, and directly creates the Skill) and Review & Improve (send an existing Skill draft, and it diagnoses issues layer by layer with improvement suggestions). When information is insufficient, it asks follow-up questions and offers choices rather than open-ended questions, reducing the effort to respond. Finally, it directly calls createSkill to create a usable Skill, no manual instruction writing needed. Ideal for creators who want to turn their experience into skills but don't know where to start.
Skill Judge
Evaluate any AI skill you want to assess—an installed skill, a marketplace page or SKILL.md text, or a spoken description of your needs. Following the principle of “test it, don’t judge by description,” Skill Judge creates a customized minimum test set with four positive, one negative, and one ambiguous use case for live verification. If the skill can be run, it is tested on the spot; otherwise, the entire process is marked 【Static evaluation】 and the findings are weighted accordingly. The output begins with a conclusion and includes a five-dimension weighted scorecard: trigger reliability (30%), execution effectiveness (25%), cost efficiency (15%), robustness and boundaries (15%), and safety and trustworthiness (15%). Overall scores are classified as ≥80: retain; 60–80: targeted refinement; <60: redo. It also identifies the causes of weaknesses—description, steps, final self-check, or length—and provides prioritized improvement suggestions. By default, you get a one-page rapid-test conclusion. Say “deep review” to upgrade to a full diagnostic report with three consecutive runs of the same task for a stability comparison, plus online verification of the source and author’s reputation. Results are automatically filed in the “Skill Evaluation Reports” group. Use it to screen skills before installing them from the marketplace, quality-check skills you create, or re-evaluate installed skills.
Garry Tan's Review Skill
Description
Recommended by
nene@YouMind
Why we love this skill
Centered on Garry Tan’s rigorous entrepreneurial and builder modes, this Skill provides CEO-level product review through six sharp questions and five generative questions, helping you validate ideas and optimize plans.
Based on gstack's office-hours and plan-ceo-review, this Skill is adapted for the YouMind environment for product review and document generation.
Related Skills
View allSKILLS Refinement Workshop
Seven processes, one closed loop. SkillForge Refinement Workshop transforms the gstack engineering sprint methodology open-sourced by YC leader Garry Tan into a full lifecycle management system for Claude Code Skills. From six soul-searching questions to verify demand authenticity, to architecture planning, SKILL.md forging, quality review, eval testing, release packaging, and finally review and iteration—each stage has dedicated AI roles, with confidence scores and anti-spam mechanisms. It also includes two additional processes: emergency diagnosis and security audit. No 'good ideas,' only evidence-based judgments. Every skill is tempered through seven processes.
Skill Designer
Can't write a Skill? Just describe it in a sentence. Skill Designer is a 'Skill that teaches you to write Skills' — it uses a 7-layer meta-framework (Intent Definition → Input Classification → Workflow Design → Tool Strategy → Output Specification → Constraints and Boundaries → Composability & Testing) to turn your vague idea into a well-structured, immediately executable YouMind Skill. It supports two paths: Design from Scratch (describe your idea, it automatically asks for missing info, runs through the 7 layers, and directly creates the Skill) and Review & Improve (send an existing Skill draft, and it diagnoses issues layer by layer with improvement suggestions). When information is insufficient, it asks follow-up questions and offers choices rather than open-ended questions, reducing the effort to respond. Finally, it directly calls createSkill to create a usable Skill, no manual instruction writing needed. Ideal for creators who want to turn their experience into skills but don't know where to start.
Skill Judge
Evaluate any AI skill you want to assess—an installed skill, a marketplace page or SKILL.md text, or a spoken description of your needs. Following the principle of “test it, don’t judge by description,” Skill Judge creates a customized minimum test set with four positive, one negative, and one ambiguous use case for live verification. If the skill can be run, it is tested on the spot; otherwise, the entire process is marked 【Static evaluation】 and the findings are weighted accordingly. The output begins with a conclusion and includes a five-dimension weighted scorecard: trigger reliability (30%), execution effectiveness (25%), cost efficiency (15%), robustness and boundaries (15%), and safety and trustworthiness (15%). Overall scores are classified as ≥80: retain; 60–80: targeted refinement; <60: redo. It also identifies the causes of weaknesses—description, steps, final self-check, or length—and provides prioritized improvement suggestions. By default, you get a one-page rapid-test conclusion. Say “deep review” to upgrade to a full diagnostic report with three consecutive runs of the same task for a stability comparison, plus online verification of the source and author’s reputation. Results are automatically filed in the “Skill Evaluation Reports” group. Use it to screen skills before installing them from the marketplace, quality-check skills you create, or re-evaluate installed skills.
Find your next favorite skill
Explore more curated AI skills for research, creation, and everyday work.