تجاوز إلى المحتوى
جرّب المرجع الدورة المحفوظات
العربية
← كل playbooks الأفراد والموارد البشرية

playbook

Run a hire from open to offer

Run a whole role end to end — inclusive JD, structured interview kit and scorecard, a fairly synthesized panel, a bias check, and the offer and rejection comms — one consistent, defensible system where the human always makes the call.

متقدّم ~half a day

متى تلجأ إلى هذا

Hiring is where every People & HR skill gets tested at once: a fair posting, a structured process, an honest verdict, and humane comms — usually improvised under deadline, with the consistency that makes a hire defensible left to chance. This is the system that runs the whole arc: open the role inclusively, interview everyone against the same scorecard, synthesize the panel fairly, check for bias, and draft both the offer and the rejections — so the process is consistent end to end and the human always makes the call. It ties the other five People & HR playbooks into one workflow.

جهّز هذا أولًا

  • The role brief and genuine must-have skills in role-notes.md — the raw input to the *Write a job description that widens the pool* playbook, which is step one here.
  • Your interview panel and who assesses what, plus a workspace your company has approved for people data — every candidate detail stays inside it.
  • Your house constraints: comp band and approvals, anything legally required in postings or rejections for your region, and what you can and can't promise in an offer.

الـ workflow

  1. Open the role inclusively

    The whole system inherits this step. Run the inclusive-JD flow so the posting screens for genuine must-haves, not a wishlist — because the skills you post become the skills you'll score against, and a biased posting biases everything downstream.

    أنت تطلب 
    Read role-notes.md. Run our inclusive-JD process: split genuine must-haves from nice-to-haves and challenge anything over-specified, draft the job description from the trimmed must-haves, scan it for language that narrows the pool (gendered/coded words, too-high experience bar, jargon, 'culture fit'), and add an inclusive close. Give me the JD plus the confirmed must-have list — I'll reuse that list to build the scorecard.

    ما تحصل عليه A posted-ready JD with a short, honest requirements section, a bias scan you've reviewed, and — crucially — a clean confirmed must-have list that becomes the spine of the interview kit.

    Keep the confirmed must-have list. Every later step — questions, scorecard, synthesis — anchors to it, which is what makes the whole process consistent.

  2. Build a structured interview kit and scorecard

    Structure is what makes hiring fair and comparable. Turn each must-have into the same questions and the same rating scale for every candidate, so you're comparing evidence against a rubric, not comparing gut feelings.

    أنت تطلب 
    Using the confirmed must-have list, build a structured interview kit: for each must-have skill, 2 behavioral questions and what a strong vs weak answer sounds like. Then a 1–4 scorecard with one row per must-have and an evidence box per row. Split the questions across the panel so each interviewer owns specific skills with no overlap, and add a note that ratings need written evidence, not just a number.

    ما تحصل عليه A question bank mapped to must-haves with strong/weak answer guides, a per-skill 1–4 scorecard with evidence boxes, and an assignment grid — so every candidate is assessed on the same things and you can synthesize cleanly later.

  3. Synthesize the panel fairly and check for bias

    After interviews, run the interview-synthesis flow: weight evidence equally, keep real disagreements visible, and flag 'culture fit' language as its own pass — turning scattered scorecards into a fair, defensible read.

    أنت تطلب 
    Here are the five completed scorecards and notes for [candidate]. Synthesize against the rubric: strengths and concerns corroborated across interviewers (weight evidence equally, NOT by who wrote most), the genuine disagreements with both sides quoted and a tiebreaker for each, and a separate pass flagging any bias or 'culture fit' language that isn't job-related. Map the candidate to each must-have. Do NOT recommend hire or no-hire.

    ما تحصل عليه A neutral, scorecard-anchored brief per candidate — corroborated strengths and concerns, the real splits to resolve, bias flags to set aside — with no verdict baked in, ready for the panel to decide.

    Run this identically for every finalist. Same rubric, same synthesis, same bias check — consistency across candidates is what makes a decision defensible, not just thorough for one person.

  4. Let the humans decide, then draft the offer

    The panel makes the call against the rubric; only then does Claude draft. Run the offer with the same care as the *Draft a sensitive message you can stand behind* playbook — warm, clear, no promise outside the approved band — with the real terms as placeholders.

    أنت تطلب 
    The panel has decided to extend an offer. Draft a warm, clear offer message for the [role]: enthusiasm without over-promising, the key terms as [placeholders] (title, comp, start date, manager), a clear acceptance step, and a contact for questions. Keep it honest about what's confirmed vs pending final paperwork. I'll fill the placeholders from our approved system.

    ما تحصل عليه A warm, honest offer draft with every real term as a [placeholder] you fill from your HR system — no comp number or promise that wasn't approved, and a clear, kind path to yes.

    The decision is the panel's, made against the rubric — Claude drafts the message, never the verdict. Fill comp and terms only from your approved system, never invented in chat.

  5. Close the loop with humane rejections

    How you reject is your employer brand. Use the *Draft a sensitive message you can stand behind* approach to write respectful, honest declines that keep names and unsharable feedback out — closing the role consistently for everyone, not just the person who got the offer.

    أنت تطلب 
    Draft rejection messages for the candidates we're not moving forward with: respectful, genuinely appreciative of their time, honest but kind, and consistent in tone with the offer. Keep it short. Do NOT include private panel notes or comparative feedback unless it's safe and accurate to share — use a [placeholder] where specific feedback could go, and flag where I should check before sending any.

    ما تحصل عليه Short, respectful rejection drafts consistent with the offer's warmth, free of private notes, with [placeholders] and check-before-send flags wherever specific feedback might be shared — so every candidate gets a clean, humane close.

    Read every rejection word-for-word before it sends. A careless decline undoes a careful process — and unverified feedback can create real risk, so flag anything you're unsure about.

اجعله ملكك

  • **Reuse the kit:** save the JD → interview kit → synthesis → comms flow as a /hire custom command or skill (see the Playbook's *Features* tab) so every hiring manager runs the same structured, bias-checked process instead of improvising their own.
  • **Onboard the winner:** the moment the offer is accepted, hand off to the *Build a reusable onboarding-in-a-box* playbook — the must-have list you wrote here becomes the 30-day ramp goal, so hiring and onboarding share one source of truth.
  • **Bulk roles:** for high-volume hiring, keep one JD and one scorecard per role and run the *Turn five interviewers into one fair verdict* synthesis per candidate — the rubric stays fixed, only the candidates change, which keeps a big funnel fair.
  • **Plain-English offer pack:** before the offer goes out, run the offer letter and any policy boilerplate the candidate has to agree to (PTO, equity, remote-work rules) through the *Rewrite a policy into plain English* playbook — so they understand exactly what they're saying yes to, not a wall of legalese.

انتبه إلى

  • Every candidate detail stays in a workspace your company has approved for people data — names, notes, comp, and decisions never go into a tool you don't trust. Confidentiality is the whole job here, not a footnote.
  • Verify the load-bearing facts at every stage: that the bias scan was actually read, that scorecards carry written evidence not just numbers, and that comp and offer terms come from your approved system — Claude will state a number confidently whether or not it's the right one.
  • Claude drafts the JD, the kit, the synthesis, and the comms — but the human always makes the call. Hire, reject, and offer decisions are the panel's, and anything touching hiring law, comp, or protected characteristics goes past a real person before it ships.

ستحصل في النهاية على A consistent, defensible hire from open to offer in one workflow — an inclusive posting, a structured scorecard everyone is judged against, a fairly synthesized and bias-checked panel, and humane offer and rejection comms — with the human owning every decision and Claude doing the drafting.