Writing Great Inputs
The quality of your generated skill depends directly on the quality of your input. SkillThis assesses your input on methodology specificity, scoring it from 0 to 100.
What Makes a High-Scoring Input
Section titled “What Makes a High-Scoring Input”The AI evaluates your input for concrete, actionable methodology. Here’s what it looks for:
| Element | Example | Impact |
|---|---|---|
| Step-by-step processes | ”First I do X, then Y, then Z” | High |
| Named tools and frameworks | ”I use the MATCH framework, LinkedIn Recruiter” | High |
| Decision trees | ”If X happens, I do Y; otherwise Z” | High |
| Quantifiable outcomes | ”Screened 50, interviewed 15, placed 3” | High |
| Clear input/output | ”Given a job req, I produce a candidate shortlist” | Medium |
| Domain-specific vocabulary | ”Boolean searches”, “RICE scoring”, “runbooks” | Medium |
Scoring Tiers
Section titled “Scoring Tiers”High Score (70-100)
Section titled “High Score (70-100)”Your input contains concrete methodology. Skills generated from these inputs typically grade B or higher.
I'm a technical recruiter specializing in engineering roles atearly-stage startups. My process: I source candidates throughLinkedIn boolean searches and GitHub activity, screen them withmy MATCH framework (Motivation, Aptitude, Technical depth,Culture fit, Hunger), then guide them through the full interviewloop. I always prep candidates before each round and debriefhiring managers after.This scores high because it names specific tools (LinkedIn, GitHub), a named framework (MATCH), a clear process flow, and concrete actions.
Medium Score (40-70)
Section titled “Medium Score (40-70)”Some methodology but incomplete. SkillThis will ask extraction questions to fill gaps.
Program manager focused on product launches. I coordinateengineering, design, marketing, and legal. I'm good at identifyingdependencies early and keeping stakeholders aligned withoutendless meetings.This mentions coordination and dependencies but lacks specific tools, frameworks, or step-by-step processes.
Low Score (0-40)
Section titled “Low Score (0-40)”Abstract or aspirational. SkillThis will prompt you for much more detail.
tech recruiter, 8 years exp. i source, screen, and closeengineering candidates. good at passive outreach and salarynegotiation.This lists capabilities without explaining how. No tools, no process, no examples.
What to Include
Section titled “What to Include”A strong input covers at least 3 of these:
- Your process - What do you do first, second, third?
- Your tools - What specific software, frameworks, or techniques do you use?
- Your criteria - How do you evaluate quality or make decisions?
- A real example - Walk through an actual case from start to finish
- Your edge cases - What do you do when things go wrong?
What to Avoid
Section titled “What to Avoid”| Anti-Pattern | Example | Why It Fails |
|---|---|---|
| Vague goals | ”I design revolutionary UIs” | No methodology |
| Identity statements | ”I’m a great engineer” | No process |
| Lists without depth | ”Python, JavaScript, React” | Tools without workflow |
| Philosophy without action | ”I believe in user-centered design” | Principles without steps |
| Too short | ”marketing” | Nothing to work with |
Side-by-Side Comparison
Section titled “Side-by-Side Comparison”Sales Engineer
Section titled “Sales Engineer”Brief input (low score):
SE at a devtools company. I do demos, POCs, and help closetechnical deals. good at translating features into business valueDetailed input (high score):
I'm a sales engineer who runs technical demos and POCs forenterprise SaaS deals. My approach: I never demo features, I demooutcomes. Before any call I research the prospect's tech stack,identify their likely pain points, and build a custom demoenvironment that mirrors their setup. During the call I askdiscovery questions first, then show only the 3-4 features thatsolve THEIR problems. I always leave time for hands-on exploration.Post-demo I send a personalized follow-up with a Loom walkthroughof their specific use case.The detailed version names specific techniques (custom demo environments, discovery questions, Loom walkthroughs), describes a clear process (research, build, present, follow up), and includes decision criteria (only show features that solve their problems).