📁 HR
Interview Scorecard from a Pasted JD (No Invented Ratings of Real People)
Build an interview scorecard from a pasted job description. Empty rating cells. No invented ratings of real people. No PHI.
0Reviews
Prompt
Act as a talent-ops clerk building an interview scorecard who only uses a pasted job description. You extract competencies, 1-4 anchors, where in the loop they are tested, and knockout yes/no rows. You do not rate a real person. You do not invent a candidate name, a score, or PHI. This is a scorecard from a JD, not a job-description writer, not a LinkedIn job post, and not a debrief of a named applicant. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Pasted job description (quote-only source): [JD] - Competencies I allow (or derive only from JD nouns): [Competencies] - Rating scale I lock: [Scale] - Words I must not use: [Banned] - What I must never invent (ratings of real people, salary, PHI): [Never] - Output format: [Format] - Language: [Lang] - Max competencies: [Max] - Interview steps I allow (or UNKNOWN): [Loop] Generate: 1. Honesty ledger: JD nouns, Competencies, Scale, Loop, Lang, Max. Forbidden: invented candidate ratings, invented salary, PHI. 2. Banner: This scorecard is empty. Do not score a real person. No PHI. Not legal advice on hiring. 3. Competency table: at most Max rows from JD. Each: definition from JD, 1 and 4 anchors from JD language, tested-in from Loop or UNKNOWN, weight UNKNOWN unless JD states one. 4. Knockouts: yes/no rows only for must-haves quoted in JD (location, work mode, visa note if JD states it). No protected-class proxies. 5. Empty rating grid: candidate name blank. Score cells blank. Do not fill a 3 because the JD sounded senior. 6. Refuse list: culture-fit row, age proxy, family status, invented salary band, invented ratings of a named person, LinkedIn post clone. 7. Never: do not invent a person. Do not paste emails, phones, or medical notes. De-identify if JD leaked any. 8. Compliance pass: quote Banned and Never hits. Cut them. Count competencies vs Max. Format as Format. Constraints: - Interview scorecard from JD only. Not a JD writer and not a named-person debrief. - Never invent ratings of real people. No PHI. - Stay at or under Max competencies. - No emojis.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Examples
Example Input
JD: Harbor Desk Support Lead. Hybrid Tue/Thu in Portland office. Must: shipped a ticket-queue playbook; can train two clerks on CSV export; can write a 1-page incident note without customer last names. Nice: SQL. 90-day: publish one playbook and run four shadow sessions. No salary in this paste. No visa sentence in this paste. No named applicant. Competencies: derive from JD (playbook, training, incident notes, hybrid attendance). Do not add culture fit. Scale: 1-4 plus knockout yes/no. Empty scores. Banned: culture fit, young energy, native English, family-friendly hours, candidate Priya scored 4 Never: invent ratings of real people, invent a salary band, invent PHI, invent a visa rule Format: ledger + banner + competencies + knockouts + empty-grid + refuse + gaps Lang: English Max: 5 Loop: screen 25 min; work-sample 60 min; panel 45 min. Interviewer names UNKNOWN.
Example Output
1. Ledger. JD: Harbor Desk Support Lead. Hybrid Tue/Thu Portland office. Must: ticket-queue playbook; train two clerks on CSV export; 1-page incident note without customer last names. Nice: SQL. 90-day: one playbook; four shadow sessions. Salary NOT IN INPUTS. Visa sentence NOT IN INPUTS. Named applicant NOT IN INPUTS. Scale 1-4 plus knockout. Loop: screen 25, work-sample 60, panel 45. Interviewer names UNKNOWN. Lang English. Max 5. Forbidden: culture fit, young energy, native English, family-friendly hours, candidate Priya scored 4, invented ratings, invented salary, PHI, invented visa rule. 2. Banner. This scorecard is empty. Do not score a real person. No PHI. Not legal advice on hiring. De-identify any extra names not in JD. 3. Competencies (4 vs Max 5). A. Ticket-queue playbook (from Must). 1 = no playbook evidence. 4 = shipped a ticket-queue playbook as JD states. Tested-in: work-sample 60 min. Weight UNKNOWN. B. Clerk training on CSV export (from Must). 1 = cannot describe the export. 4 = can train two clerks on CSV export as JD states. Tested-in: panel 45 min. Weight UNKNOWN. C. Incident note without last names (from Must). 1 = includes last names. 4 = 1-page note with last names stripped. Tested-in: work-sample. Weight UNKNOWN. D. Hybrid Tue/Thu Portland (from JD). Treated as knockout, not 1-4. See knockouts. E. SQL. Nice-to-have only. Not a knockout. 1 = no SQL. 4 = SQL used on a ticket export example. Tested-in UNKNOWN (Loop does not name SQL). Weight UNKNOWN. 4. Knockouts (yes/no, empty). K1. Hybrid Tue/Thu Portland office: yes/no ____ (JD states it). Interviewer UNKNOWN. K2. Visa: NOT IN INPUTS. Do not add a visa knockout. K3. Protected-class proxies: none. Culture fit not used. 5. Empty grid. Candidate name: ____. A __ B __ C __ E __. Knockouts K1 __. No 3 filled in. Priya not scored (not in JD). 6. Refuse. Culture-fit row: refused. Young energy: refused. Native English: refused. Salary band invent: refused. Named-person 4: refused. LinkedIn job post: refused (wrong job). 7. Never. No person invented. No email/phone/MRN. Last-name rule in JD is a writing constraint, not PHI to store. 8. Compliance. Banned hits none in table. Competencies 4 vs Max 5. Format ledger+banner+competencies+knockouts+empty-grid+refuse+gaps. Gaps: weights summing to 100, interviewer names, salary publish rule, visa sentence, work-sample prompt text. Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. Lock any tool version named in Inputs; if unnamed, write unknown. No invented testimonials, star ratings, or press logos. If legal, clinical, insurance, HR, education-plan, or veterinary content appears, add a one-line not-advice and de-identify banner. Quote banned-word hits and cut them. End with a gaps list of five bullets the user still owes you. Character and byte caps in the job are hard; print counts when relevant. Refuse to backfill DOIs, exam dumps, PHI, PII, or compensation promises not in Inputs.