The AI Interview Deconstructor

** You submitted forty-three applications last quarter. You know your craft cold — you've shipped real products, solved real problems, and can hold your own in any technical conversation. But the callbacks aren't coming. Not from the roles you actually want. The rejections land at 2am, automated, no explanation. You've started wondering if something is wrong with you. Nothing is wrong with you. Something is wrong with your answers — specifically, how they read to a machine that never sees your GitHub, your track record, or the look on a client's face when you solved their problem. AI screening systems score language patterns, sentence structure, keyword density, and response framing before a human ever opens your file. You're being evaluated by a rubric you've never seen, optimized for a reader you've never met. The obvious fix — tweaking your wording, studying common interview questions, asking a friend to review your answers — doesn't touch the actual problem. Your friend is a human. They're optimizing for authenticity and relatability. The bot is optimizing for signal phrases, structural completeness, and tone consistency. Those are different games with different rules. This skill gives you a structured system for reverse-engineering those rules. It analyzes your responses against the scoring dimensions AI screening algorithms actually use, then rebuilds your answers from the inside out — same substance, language the algorithm registers as high-signal. What you get: SKILL.md — the core analysis protocol, including response evaluation criteria, rewrite sequences, and output specifications for AI-optimized interview answers. SETUP_GUIDE.md — step-by-step instructions for deploying this skill in your agent environment and calibrating it to your target role type. EXAMPLES.md — before/after response pairs across common interview question categories, showing exactly what changes and why. CHANGELOG.md — version history documenting methodology updates as screening algorithm behavior evolves. External dependency: None. Runs entirely within your agent's context. --- CORE CAPABILITIES: - Deconstruct your existing interview responses against known AI screening evaluation dimensions - Identify language patterns, structural gaps, and keyword absences that trigger low-confidence scores - Rebuild answers to preserve your authentic experience while optimizing for algorithmic readability - Score responses across tone consistency, structural completeness, and role-signal alignment in a single pass - Translate technical accomplishments into language that registers clearly across both bot and human reviewers - Calibrate output framing to match the seniority level and domain of your target role - Generate side-by-side comparisons of original and optimized responses with change rationale - Adapt the evaluation protocol to specific job descriptions so every answer targets the exact role you're pursuing