@autostackup/hr

The only HR MCP you'll ever need.

Professionally built candidate screening, 30/60/90-day onboarding plans, offer letter generation, interview kits, and performance reviews — frameworks used by top People teams, all inside Claude. What HR consultants charge thousands for, you get in one prompt.

Tools

screen_candidate

Score a candidate against your role's required skills, experience level, and must-haves. Uses structured competency scoring across four dimensions: Skills Match (35%), Experience Level (30%), Background Quality (25%), Red Flag Risk (10%). Hard-disqualifies on any must-have miss.

Returns: Tier (A/B/C/D), overall score, dimension scores, skill gaps, salary fit assessment, recommended next step.

Try it: "Screen this candidate for a Senior Engineer role: 5 years backend Python, AWS experience, no system design background, asking $180k"

draft_offer_letter

Generate a complete, ready-to-send offer letter with full compensation summary, equity and bonus clauses, offer expiry date, negotiation talking points, and a follow-up call script. Supports formal, warm, and startup tones.

Returns: Full offer letter text, compensation summary table, negotiation notes, follow-up script.

Try it: "Draft an offer letter for a Senior Engineer, $160k base, 0.5% equity, $10k signing, warm tone, expires in 5 days"

create_onboarding_plan

Build a complete 30/60/90-day onboarding plan for any new hire. Includes Day 1 checklist, weekly themes and goals, milestone definitions, manager action checklist, red flag warnings, and a success definition. Adapts for individual contributor, manager, and executive roles.

Returns: Day 1 checklist, 30/60/90-day plans with weekly breakdown, manager guide, success metrics.

Try it: "Create a 30/60/90 onboarding plan for a new Head of Marketing joining a 30-person SaaS startup"

build_interview_kit

Generate a complete interview kit for any role and interview type. Structured STAR questions by competency, follow-up probes, "what good looks like" benchmarks, red flag questions, a scorecard template, legal reminders, and a debrief facilitation guide.

Supports: Screening, technical, behavioural, case, panel, and final interviews.

Returns: Question bank by competency, follow-up probes, scorecard, legal reminders, debrief guide.

Try it: "Build a behavioural interview kit for a VP of Sales role, 5 competencies, 60-minute format"

write_performance_review

Write a structured performance review with achievements narrative, areas for development, next-period goals, compensation recommendation, and a manager meeting prep guide. Adapts to all rating levels and supports coaching, direct, and formal tones.

Ratings supported: Exceeds expectations / Meets expectations / Partially meets / Does not meet

Returns: Full review narrative, development plan, goals section, comp recommendation, manager meeting script.

Try it: "Write a performance review for a Product Manager, meets expectations, strong on delivery but needs to improve stakeholder communication, eligible for 5% merit increase"

Setup

{
  "mcpServers": {
    "autostackup-hr": {
      "command": "npx",
      "args": ["-y", "@autostackup/hr"]
    }
  }
}

No API keys required. All tools run entirely on Claude.

Part of the Autostackup Suite

| Package | Description | |---------|-------------| | @autostackup/founder | All tools in one server + founder orchestration | | @autostackup/sales | BANT/MEDDIC qualification, ICP scoring, outreach | | @autostackup/marketing | Kotler audit, IMC planning, audience segmentation | | @autostackup/hr | ← You are here |

License

MIT — Autostackup