From interview
Generate a pre-screening assessment for a candidate from their resume and the job description. Produces a structured strengths/gaps analysis and two targeted interview questions. Use before any interview to prepare. Saves to questions/<CandidateName>_assessment.md.
How this command is triggered — by the user, by Claude, or both
Slash command
/interview:assess-candidate <Candidate Name> [--resume <path>] [--jd <path>]The summary Claude sees in its command listing — used to decide when to auto-load this command
# Pre-Screening Candidate Assessment Produce a structured pre-screening assessment from a candidate's resume before any interview takes place. ## Usage ## Project Structure ## Step 1: Discover Files If paths are not provided, locate them: - **Resume**: search `candidates/` for a PDF whose filename contains the candidate's name (case-insensitive, partial match) - **Job description**: use the first PDF found in `job-description/` - **Interviewer context**: read `dnuckolls_intro.md` if it exists — use it to ground the assessment in the interviewer's priorities and the team's current ...
Produce a structured pre-screening assessment from a candidate's resume before any interview takes place.
/interview:assess-candidate <Candidate Name>
/interview:assess-candidate <Candidate Name> [--resume <path>] [--jd <path>]
<project-root>/
├── candidates/ # Resumes (PDF), matched by candidate name
├── questions/ # Output location for _assessment.md files
├── job-description/ # Job description PDF (auto-detected, first PDF found)
└── dnuckolls_intro.md # Interviewer background and priorities
If paths are not provided, locate them:
candidates/ for a PDF whose filename contains the candidate's name (case-insensitive, partial match)job-description/dnuckolls_intro.md if it exists — use it to ground the assessment in the interviewer's priorities and the team's current stateReport which files you found before proceeding.
Read the job description, resume, and interviewer intro in full. Extract:
Write a markdown assessment to questions/<CandidateLastName><CandidateFirstName>_assessment.md (no spaces, last name first).
Use exactly this structure:
# Candidate Assessment: <First Last>
**Role:** <Role Title from JD>
**Experience:** <total years, key domains>
---
Bullet list. For each item: name the specific signal, tie it to a JD requirement. Do not pad — only include real signals.
Bullet list. For each item: name what is absent or uncertain, explain why it matters for this role. Distinguish "confirmed absent" from "not evidenced."
One paragraph. Lead with a direct signal/risk statement — e.g., "Strong on X; significant uncertainty on Y." Name the central tension if one exists. Be honest about unknowns; don't oversell.
Choose the two areas of highest uncertainty or highest importance from the gaps analysis. Write one targeted question per area.
<Area being probed>> "<exact question — specific to this candidate's background, not generic>"
What to listen for: Two to four sentences. Name what a strong answer includes, and what a red flag looks like. Make the distinction concrete enough that a different interviewer could use it.
<Area being probed>> "<exact question>"
What to listen for: Two to four sentences.
The assessment should read like a memo from a senior recruiter to the hiring manager — specific, honest, and useful as a preparation tool. Avoid praise inflation. A candidate with genuine strengths and real gaps should look exactly that way on paper.
The questions should be tailored to this candidate's resume — they should not work as generic interview questions. Reference the candidate's actual projects, companies, or stated experience when possible.
Report the path to the written assessment and give a one-sentence summary of the overall fit verdict.
npx claudepluginhub glorykidd/marketplace --plugin interview