CV Scanner Skill

The problem

Manual resume screening is prone to unconscious bias and inconsistent grading. Processing high volumes of CVs while maintaining an audit trail for compliance is slow and prone to human error.

What it does

• Anonymizes CVs by stripping protected identity characteristics before the scoring pass.
• Extracts and validates candidate evidence against a structured, weighted scoring rubric.
• Calculates confirmed score bounds and identifies candidates requiring manual oversight due to OCR quality or missing data.
• Generates a spreadsheet-ready audit package including evidence logs, result CSVs, and a run manifest.
• Ranks candidates based on verified evidence while flagging must-have gaps for human review.

Frameworks & tools

Python scripts for extraction and validation, YAML for rubric schemas, and CSV/JSON for auditable outputs.

Why this beats prompting it yourself

Simple prompts often hallucinate candidate details or fail to maintain strict anonymity. This skill enforces a structured pipeline that separates identity from evidence, uses deterministic IDs, and provides a multi-step audit trail that ad-hoc prompts lack.

Use cases

• Initial blind screening for high-volume recruitment cycles.
• Compliance-focused hiring where scoring must be mapped to specific CV locators.
• Standardizing candidate evaluation across multiple hiring managers using a shared rubric.
• Auditing recruitment agency submissions against internal role requirements.

Known limitations

Requires an approved rubric and job description to begin. Cannot make final hiring decisions or send automated rejections.