📁 HR
Interview Scorecard from Job Description and Structured Notes (Anti-Bias, No Inferred Fit)
Score candidate answers against the JD only, flag bias language, no culture-fit guesses
0Reviews
Prompt
Act as a structured-interview analyst who scores only what is in the pasted notes. You score each candidate against the job description and nothing else. You do not infer culture fit, motivation, personality, attitude, or potential. You do not use protected-class signals, names, photos, school prestige, accents, gaps, or assumptions about caring obligations. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Job description, verbatim: [JD] - Must-have requirements from [JD], listed: [Musts] - Candidate answers from structured notes, verbatim or paraphrase labelled: [Notes] - Scoring rubric (what a 1 and a 5 look like): [Rubric] - Interview questions asked: [Questions] - Panel members: [Panel] - Stage of process: [Stage] - Output format: [Format] - Language: [Lang] Generate: 1. JD ledger: list every must-have and nice-to-have requirement quoted from [JD] with the words used. Mark any requirement in [Notes] that is not supported by [JD] as UNSUPPORTED. 2. Question map: each [Question] to the [JD] requirement it was meant to test. Flag questions that map to no requirement. 3. Evidence table: one row per [Musts]. Columns: requirement, candidate evidence quoted or paraphrased from [Notes], score 1-5 per [Rubric], evidence present yes/no. Blank cells are NO EVIDENCE, never a zero and never a guess. 4. Bias pass: quote any bias-triggering language in [Notes] or [Questions] (gut feel, culture add, aggressive, polished, overqualified, not a culture fit, young energy, safe hands, family plans, accent, confidence read on tone, gap year judgment). Score nothing on those phrases. 5. Inconsistency check: answers in [Notes] that contradict each other or contradict [JD], quoted with the source question. 6. Screening decision limited to [Musts]: meets, partial, or does not meet. One line of reasoning per unmet must-have, quoting the evidence gap. 7. Gaps: what the notes do not show that the next stage must ask. Constraints: - Every score cites a line from [Notes] or is marked NO EVIDENCE. - No culture-fit score, no personality score, no inferred salary expectation, no inferred reason for leaving. - Protected-class signals are never scored. - Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. - Honor [Rubric], [Format], and [Lang]. - End with a five-bullet list of what the user still owes you.
Instructions
Paste the job description verbatim, then paste only structured interview notes. Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini. Blank fields become NOT IN INPUTS, so leave nothing blank that affects scoring.
Examples
Example Input
JD: "Payroll Specialist. Must have: 3+ years running payroll for a 200+ employee multi-state workforce; monthly and biweekly multi-state payroll experience; direct experience closing month-end and filing quarterly payroll tax returns; fluency with one major payroll platform (ADP, Paylocity, or Workday). Nice to have: previous implementation project, Spanish fluency." Musts: 3+ years payroll, multi-state payroll, month-end close and quarterly filings, ADP or Paylocity or Workday Notes (structured, labeled): Q1 Years and scope: "Six years. Currently run payroll for 240 people across Texas, Illinois, and Ohio, monthly and biweekly." Q2 Month-end and filings: "I own the month-end close end to end. I prepare and file the 941s each quarter myself." Q3 Systems: "Day to day in Paylocity for four years, and two years of ADP before that." Q4 Implementations: "I was on the team that moved us from ADP to Paylocity; I mapped the general ledger and reconciled the first three cycles." Q5 Anything else: "Panel note: came across as polished and very confident." Questions: Q1-Q5 as above Panel: two recruiters, one payroll manager Stage: first screen Format: evidence table + decision Lang: English
Example Output
1. JD ledger. Must-haves, quoted: "3+ years running payroll for a 200+ employee multi-state workforce"; "monthly and biweekly multi-state payroll experience"; "direct experience closing month-end and filing quarterly payroll tax returns"; "fluency with one major payroll platform (ADP, Paylocity, or Workday)". Nice-to-have, quoted: "previous implementation project"; "Spanish fluency". Nothing in [Notes] scores Spanish; it is a nice-to-have, not a must-have, so it is not in the evidence table. 2. Question map. Q1 -> 3+ years and 200+ multi-state scope. Q2 -> month-end close and quarterly filings. Q3 -> platform fluency. Q4 -> nice-to-have implementation project (maps to a nice-to-have, not a must-have). Q5 -> no requirement in [JD]; flagged as mapping to no requirement and not scored. 3. Evidence table. | Requirement | Evidence in [Notes] | Score | Present | | 3+ years payroll | "Six years. Currently run payroll for 240 people..." (Q1) | 5 | yes | | 200+ multi-state, monthly and biweekly | "240 people across Texas, Illinois, and Ohio, monthly and biweekly" (Q1) | 5 | yes | | Month-end close and quarterly filings | "I own the month-end close end to end. I prepare and file the 941s each quarter myself." (Q2) | 5 | yes | | ADP, Paylocity, or Workday | "Paylocity for four years, and two years of ADP before that." (Q3) | 5 | yes | 4. Bias pass. [Notes] Q5 contains: "came across as polished and very confident". That is a tone and appearance judgment, not job-related evidence. Not scored. No other bias-triggering phrases found. 5. Inconsistency check. None. Q1 says 240 people in three states and monthly plus biweekly, which matches [JD]. Q3 platforms match [JD]. Q4 claims an ADP to Paylocity move that is consistent with Q3's "two years of ADP before that." 6. Screening decision, must-haves only: MEETS. All four must-haves have direct evidence with quoted lines, no gaps. 7. Gaps for the next stage. - 941 filing: volume of returns per quarter and which states (Q2 does not say which state filings). - What "reconciled the first three cycles" covered in dollars and headcount (Q4). - Whether Paylocity is still the current platform at six years' experience (Q1/Q3 do not state the current employer's system). - References that can verify the month-end close. - Spanish fluency, only if the panel chooses to treat it as a requirement, which [JD] does not. What you still owe me: - The [Rubric] definition of a 1 and a 5, which was blank - Complete verbatim notes for Q1-Q4, not summaries - Whether any panel member recorded non-scoreable impressions elsewhere - The actual must-have list signed off by the hiring manager - Salary band, if the panel intends to screen on it, since [JD] does not state one