Evidence-Based Resume Review

Create an auditable first-pass review package from a structured job description and authorized resume text.

Three-Step Workflow

1. 1. Define only job-related requirements in job.json, including must-have skills, optional skills, weights, aliases, and minimum relevant experience.
2. Run the local reviewer or import the n8n workflow. The pipeline removes direct contact fields and protected-trait phrases before matching requirements to quoted resume evidence.
3. Give the generated queue, evidence report, and interview questions to an authorized human reviewer. The workflow never produces an automatic hire or reject decision.

Quick Start

CODEBLOCK0

Outputs:

• - screening_report.json: per-requirement evidence and scoring details.
• INLINECODE2: sortable review order using anonymous candidate IDs.
• INLINECODE3: evidence-based questions for a human interviewer.

Non-Negotiable Rules

• - Process only resumes the organization is authorized to review.
• Do not use name, phone, email, photo, gender, age, birth date, ethnicity, marital status, health, disability, religion, political affiliation, or other protected traits for scoring.
• Treat a missing keyword as missing evidence, not proof that the candidate lacks the skill.
• Keep all candidates in human_review_required; never auto-send acceptance or rejection messages.
• Store source resumes separately with restricted access and a documented retention period.
• Validate local employment, privacy, and anti-discrimination requirements before production use.

Read references/usage.md for schemas and recovery steps. Read references/fairness-and-privacy.md before connecting real candidate data.