Personalized rejection feedback with Claude

A Claude Skill that takes a rejected candidate’s interview scorecards and (where available) interview transcripts, drafts a specific, useful rejection email or recruiter-call talking points, and produces the recruiter-side notes for the call. Replaces the form-letter rejection that damages candidate experience with personalized feedback the candidate can actually use.

What you’ll need

• Claude Code or Claude.ai with custom Skills enabled and Tier-A authorization for candidate data per AI policy
• Scorecard data export from the ATS — Ashby, Greenhouse, or Lever
• Optional: interview transcripts from BrightHire or Metaview for richer feedback grounding
• The role’s structured-interview rubric

Setup

1. Drop the Skill. Place rejection-feedback.skill into your Claude Code skills directory. The Skill exposes two callable functions: draft_email and draft_call_notes.
2. Configure the rubric. The Skill reads role rubrics from rubrics/{role_id}.yaml — same source the interview debrief Skill uses.
3. Configure the legal/HR review. Some jurisdictions limit what feedback can be shared with rejected candidates; verify your firm’s HR-counsel guidance and configure the Skill’s content rules accordingly.
4. Test on closed candidates. Run on candidates already rejected; compare drafts to what the team actually sent. Tune the prompt for tone and specificity.

How it works

For each rejected candidate, the Skill:

1. Reviews the interview record. Pulls scorecards across all interviewers, identifies the dimensions where the candidate scored highest and lowest, identifies the specific evidence that supported the rejection.
2. Filters for safely-sharable feedback. Excludes feedback that’s protected under the firm’s HR-counsel guidance (legal-risk areas, protected-class proxies, irrelevant personal observations).
3. Drafts the candidate-facing message. Specific to the candidate, grounded in interview evidence, framed constructively. Distinct from generic rejection templates.
4. Drafts recruiter-call notes (for senior or high-potential candidates getting a call rather than email). Bullet points with specific observations and suggested phrasing.

Output

Per candidate:

• Rejection email draft — opens with appreciation, summarizes what went well, identifies the specific factors that drove the decision, encourages future application where appropriate. ~150-250 words.
• Recruiter call notes (for tier-2+ candidates) — bullet points the recruiter uses on the call, including suggested responses to likely candidate questions
• Routing recommendation — email vs call vs no-response, based on candidate seniority, interview-stage reached, and pipeline-fit signal

Where it fits

Use this Skill specifically for candidates who reached Stage 3+ interviews — candidates who invested significant time in the process and deserve real feedback. Earlier-stage rejections (resume screen, recruiter screen) typically get appropriately-handled by templated rejection.

The compounding benefit: candidates rejected with real feedback are 30-50% more likely to apply again to the company in the future, more likely to refer others, and meaningfully less likely to leave damaging Glassdoor reviews.

Watch-outs

• Legal review on what feedback is shareable. Specific feedback can become evidence in employment-law claims if framed wrong. HR-counsel sign-off on the Skill’s content guardrails matters.
• Don’t auto-send. Recruiter reviews and edits every email before sending. Auto-sent rejection feedback produces inappropriate-content incidents.
• Don’t feedback on protected-class proxies. “Cultural fit” rejections, age-related rejections, family-status considerations — these damage candidates and create legal risk. The Skill’s content rules should explicitly exclude these.
• Calibrate per role and seniority. Senior-leadership rejections need different framing than entry-level rejections; configure rubric-by-role.
• Privacy. Interview content is sensitive; verify the Skill operates within Tier A enterprise AI per AI policy.