メニュー
Screen job candidates from your ATS using AI-powered evaluation. Use when asked to "screen candidates", "review applicants", "run candidate screening", "check new applications", or "evaluate candidates".
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | candidate-screening |
| description | Screen job candidates from your ATS using AI-powered evaluation. Use when asked to "screen candidates", "review applicants", "run candidate screening", "check new applications", or "evaluate candidates". |
Goal: Identify candidates we'd be thrilled to have on the team. Not "can do the job" — candidates we're excited about.
"Our first 100 people are our cultural co-founders."
These traits apply to EVERY role. A candidate missing any of these is unlikely to be great.
What it is: Takes initiative. Makes things happen. Doesn't wait for permission or perfect conditions.
How to spot it:
Red flags:
What it is: Perseveres through hard things. Doesn't quit when it gets difficult.
How to spot it:
Red flags:
What it is: Demonstrable results. Outcomes, not just activities.
How to spot it:
Red flags:
What it is: Everyone should code, build, or have deep technical understanding.
How to spot it:
Red flags:
What it is: Uses AI tools daily, understands implications, sees opportunities.
How to spot it:
Red flags:
What it is: Writes clearly, explains complex ideas simply, collaborates well.
How to spot it (their APPLICATION is a sample!):
Red flags:
What it is: Top 1% in at least one domain. Demonstrable excellence.
How to spot it:
Red flags:
What it is: Curious, diverse interests, unique perspective. Someone you'd want at dinner.
How to spot it:
Red flags:
Beyond universal traits, each role has specific requirements:
cd [YOUR_SCREENING_TOOL_PATH] && npm run screen -- --days-back 1 --min-score 70
Common variations:
# Last 7 days, specific role
npm run screen -- --days-back 7 --job-id [JOB_ID]
# Higher bar (80+), tag in ATS
npm run screen -- --min-score 80 --tag-ats
# All options
npm run screen -- --days-back 7 --min-score 75 --tag-ats --send-slack
cat [YOUR_SCREENING_TOOL_PATH]/output/daily-digest-$(date +%Y-%m-%d).md
For anyone scoring 75+, provide deeper analysis with links always included:
## [Name] — [Score]/100
**Links**: [ATS](URL) | [LinkedIn](URL or "Not provided")
### Universal Traits Assessment
| Trait | Signal | Evidence |
| ------------- | -------- | ------------------------------- |
| High Agency | ✅/⚠️/🚩 | [specific quote or observation] |
| Grit | ✅/⚠️/🚩 | [specific quote or observation] |
| Impact | ✅/⚠️/🚩 | [specific quote or observation] |
| Technical | ✅/⚠️/🚩 | [specific quote or observation] |
| AI-Native | ✅/⚠️/🚩 | [specific quote or observation] |
| Communication | ✅/⚠️/🚩 | [specific quote or observation] |
| World-Class | ✅/⚠️/🚩 | [specific quote or observation] |
| Interesting | ✅/⚠️/🚩 | [specific quote or observation] |
### Role-Specific Fit
[Assessment against role requirements]
### Why We'd Be Thrilled
- [Specific exceptional signals]
### Questions to Probe
- [Specific areas needing verification]
### Verdict
[Must Interview / Worth Exploring / Pass]
Always conclude with a summary table including clickable links:
## Summary
| Candidate | Score | Verdict | ATS | LinkedIn |
| --------- | ------- | --------- | ----------- | --------------------- |
| [Name] | [Score] | [Verdict] | [View](URL) | [Profile](URL) or N/A |
"90% confidence that this person can do a job only 10% of candidates could do."
We're not asking "can they do the job?" We're asking "would we be thrilled to work with them every day?"
When in doubt, pass. Better to miss someone good than hire someone mediocre.
Deep verification agent that finds hidden errors in code, analysis, and architecture. Assumes nothing is correct, drills deep, and verifies every claim independently.
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.
Expert advisor and thinking partner for strategic situations. Rapidly develops domain expertise, identifies knowledge gaps, and provides radically candid coaching to accelerate learning and decision-making.
Delegate complex reasoning, planning, or deep analysis tasks to OpenAI's GPT-5.2-Codex model. Use when tasks require extended thinking, multi-step planning, or when you need a second opinion on complex problems.
Daily sense-maker digest of newsletters and content feeds
Create shareable briefing documents from sessions, research, or learnings. Use when asked to "create a briefing", "write up findings", "summarize this session", "create a proposal", or when synthesizing work into a document others can consume. Produces dual-audience docs with human-readable content up top and structured AI context at bottom.