Skills package: Tailor Application v2.2

Stop guessing if your application hits the mark. This skill provides a rigorous, recruiter-side evaluation of your CV against any job description, followed by a humanized tailoring process that avoids all AI detection markers. It uses a structured, multi-reviewer approach to score your experience against job requirements, ensuring every claim is backed by evidence before a single word is rewritten.

The system handles the entire workflow from initial gap analysis to final delivery. It builds a locked shortlisting matrix, performs independent readings to eliminate bias, and generates a tailored CV and cover letter. Every deliverable undergoes a three-layer "Humanize" pass, involving automated vocabulary scans, burstiness checks, and attribution removal to ensure the final documents read as professional, human-authored text.

What it does

• Shortlisting Matrix: Builds and locks a weighted criteria matrix from any job description.
• Evidence-Based Scoring: Scores your CV through independent reviews, requiring exact quotes for every "yes" match.
• Gap Analysis: Produces a detailed job-match report highlighting exactly where your experience meets or misses the mark.
• Humanized Tailoring: Rewrites CVs and cover letters to match criteria while stripping all AI vocabulary and structural markers.
• Persistence: Maintains an Evidence Bank to store confirmed career facts and preferences for future applications.

How it works

1. Analyze: Provide a job description and your current CV to generate a locked criteria matrix.
2. Review: The system runs independent readings to score your match and identify evidence gaps.
3. Tailor: Based on the match report, the skill guides you through evidence interviews to fill gaps and refine wording.
4. Cleanse: Every document passes through the ai_text_check.py suite to ensure it passes AI detection tests.

Frameworks & tools

This skill uses Python for automated text analysis (python-docx, openpyxl) and integrates with Google Drive for persistent storage of matrices, rulebooks, and evidence banks.

Why this beats prompting it yourself

Generic prompts often produce "AI-flavored" text that triggers recruiter filters and fails to provide objective scoring. This skill separates meaning-reading from arithmetic, uses a formal reliability pilot, and enforces a mandatory multi-layer detection test that standard LLM outputs cannot match.

Use cases

• Auditing a CV against a specific high-stakes job posting to find hidden gaps.
• Generating "portal" and "email" versions of a CV that are optimized for both machines and humans.
• Maintaining a long-term evidence bank to speed up subsequent job applications.
• Running a pilot reliability test on a specific shortlisting method for recruitment teams.

Known limitations

Statistical checks for burstiness and sentence openers require at least 150 words of prose to be reliable. Does not support third-party detectors like GPTZero without user-provided access.