# People Management Prover MCP

> People Management Prover audits AI-generated HR recommendations for structural flaws. It forces your agent to validate hiring decisions against established I-O psychology criteria, anti-discrimination law (like Title VII or GDPR), and validated assessment methods. If the recommendation lacks job-related metrics or ignores adverse impact analysis, the tool flags it until you fix the underlying logic.

## Overview
- **Category:** productivity
- **Price:** Free
- **Tags:** hr, recruitment, hiring, people-management, bias, diversity, employment-law, assessment, feedback, performance, prover, reasoning

## Description

You're dealing with AI agents writing HR reports, right? They sound authoritative, but when you run them up against real legal or I-O psychology standards, they fall apart. Agents tend to recommend hiring based on vague crap like 'great fit' or 'positive energy.' The `validate_people_management` tool changes that. It forces actual rigor into every single evaluation. 

This server doesn't just critique the text; it audits your whole decision framework, demanding measurable criteria and proving legal compliance for whatever jurisdiction you operate in. You can’t skip these steps and still expect a 'good enough' conclusion.

The tool starts by enforcing job-related criteria. It makes sure you define every single evaluation point using metrics that are essential to the actual job function, not just some general skill set. You must use anchored scoring rubrics; there's no room for subjective judgment here. 

When it comes to bias auditing, you're protected. The system runs a full anti-discrimination analysis by calculating selection rates across all protected groups. Specifically, it checks against the 4/5ths rule, flagging potential adverse impact long before you file any paperwork or make an offer.

Legal compliance is another major checkpoint. You input your location, and the tool identifies the governing jurisdiction and every applicable labor statute—whether that’s US Title VII law, GDPR Article 22 requirements, or something else entirely. It tells you exactly what statutes matter right now so you don't get sued.

It also demands validated assessment methods. Vague 'gut feeling' decisions won't pass this check. The system requires evidence of predictive validity coefficients and structured scoring to prove your method actually predicts job performance, making sure your hiring logic is sound science, not just a hunch.

And for performance reviews? It forces you to generate developmental feedback that’s behavioral and criteria-referenced. You move way past simple praise or vague warnings like 'needs improvement.' The tool structures notes focused entirely on forward development, giving concrete actions the employee can take.

Basically, `validate_people_management` takes your proposed HR action—be it a hiring decision, a major performance review, or even a restructuring plan—and runs it through multiple technical checkpoints. It forces you to define measurable job criteria and assess how those criteria tie back directly to essential functions. You'll find that the tool insists on establishing clear lines of predictive validity for all assessments used in your process. If the assessment method can't prove its worth with hard data, you won't get a clean pass.

If you're handling European employee data, it will enforce checks against specific regulations like GDPR Art. 22. It verifies that any automated decision-making process respects your workers' rights to human review. When drafting performance notes, the tool ensures they aren’t just complaints; they must be behavioral and developmental.

It also handles the specifics of protected groups analysis for bias. You don't just get a pass/fail—you get detailed metrics showing selection rates across different demographics so you can pinpoint exactly where adverse impact exists. This level of detail is what separates an HR report from actionable, legally sound documentation.

Use this server when your agency needs to prove its logic. It’s designed for the moment you need to know if a recommendation holds up under intense scrutiny—the kind of scrutiny that involves statistical bias analysis and specific labor law citations.

## Tools

### validate_people_management
This tool forces your agent to build evidence-based HR decisions by checking for job criteria absence, adverse impact bias, legal compliance, validated assessment methods, and developmental feedback.

## Prompt Examples

**Prompt:** 
```
We interviewed 5 candidates for the product manager role. We liked Sarah best — great culture fit and really passionate. Let's extend the offer.
```

**Response:** 
```
CRITERIA_ABSENT — 'Culture fit' and 'passionate' are not job-related criteria — Rivera (2012) shows they correlate with interviewer similarity bias. Before selecting anyone: define the essential job functions, the KSAs required, behavioral indicators for each, and a scoring rubric. Then re-evaluate ALL candidates against the same rubric.
```

**Prompt:** 
```
Design a hiring process for senior engineers at a 200-person US tech company. Use structured behavioral interviews with STAR scoring, work sample tests, and blind resume review. Calculate adverse impact quarterly.
```

**Response:** 
```
HR_PROVEN — Structured behavioral interviews (r=0.51 per Schmidt & Hunter), work samples (r=0.54), blind resume screen mitigates name/gender bias, STAR scoring provides behavioral anchors. Title VII applies at 200 employees. Quarterly adverse impact analysis with 4/5ths rule across pipeline stages. Process is evidence-based and legally defensible.
```

**Prompt:** 
```
Write a performance review for John: 'John is a great team player and always brings positive energy. He meets expectations and should keep doing what he's doing.'
```

**Response:** 
```
FEEDBACK_EMPTY — 'Team player,' 'positive energy,' and 'meets expectations' are personality labels, not feedback. Hattie shows praise d=0.09 — near-zero impact. Replace with: specific behaviors observed, criteria they met or missed, and one concrete development action for next quarter.
```

## Capabilities

### Enforce job-related criteria
The tool forces you to define every evaluation point using measurable, essential job functions and anchored scoring rubrics.

### Audit for adverse impact
It performs bias analysis by calculating selection rates across protected groups, specifically checking against the 4/5ths rule.

### Verify legal compliance
The system identifies the governing jurisdiction and applicable labor statutes (e.g., GDPR Art. 22 or US Title VII) before finalizing a recommendation.

### Demand validated assessment methods
It requires evidence of predictive validity coefficients and structured scoring, rejecting vague 'gut feeling' decisions.

### Generate developmental feedback
The tool structures performance notes to be behavioral, criteria-referenced, and focused on forward development, moving beyond simple praise.

## Use Cases

### Interviewing for Senior Roles in the US
A TA Lead drafts, 'We need someone who is highly motivated and good at collaboration.' The agent runs this through `validate_people_management`. It rejects the vague terms ('motivated,' 'collaboration') and forces the lead to define specific behavioral indicators (e.g., 'quantifiable experience leading a cross-functional team of 5+'), making the process measurable.

### Reviewing Performance in an EU Office
An HR Manager drafts a performance review citing vague shortcomings. The agent runs it through `validate_people_management`. It immediately flags 'LEGAL_NONCOMPLIANCE' because the EU context requires specific documentation about automated decision-making (GDPR Art. 22), forcing the manager to add required legal caveats.

### Hiring for a Global Team
A company drafts a hiring plan that ignores local labor laws. The agent runs this through `validate_people_management`. It detects conflicting statutes (e.g., US Title VII vs. Brazilian CLT), forcing the user to select the correct governing law and adjust criteria accordingly.

### Revising a Hiring Process for Engineers
A team proposes using unstructured interviews, citing 'gut feel.' The agent runs this through `validate_people_management`. It cites meta-analysis (Schmidt & Hunter) and rejects the proposal, requiring the user to implement structured behavioral interviews with defined scoring rubrics.

## Benefits

- You stop making recommendations based on gut feeling. The tool demands predictive validity coefficients, ensuring your hiring process is backed by measurable evidence (Schmidt & Hunter data).
- It prevents discrimination lawsuits before they start. By running an adverse impact audit using the 4/5ths rule, you get documented mitigations for selection rates across protected groups.
- You never miss a jurisdictional requirement again. The server automatically identifies if Title VII, GDPR Art. 22, or CLT applies to your decision, forcing compliance checks.
- Feedback becomes actionable, not just nice. Instead of 'needs improvement,' the tool forces developmental guidance: specific behaviors, criteria missed, and one concrete action for next quarter (Hattie's model).
- Your entire process is defensible. The tool structures every input to trace back to an essential job function (KSA), giving you a bulletproof rationale when auditors show up.

## How It Works

The bottom line is: If you try to write an HR decision without measurable evidence and legal foresight, this tool will stop you until you build a defensible framework.

1. Feed the agent its initial HR recommendation (e.g., 'Hire Candidate X' or 'John needs improvement').
2. The `validate_people_management` tool intercepts this request and runs it through five structural audits: criteria, bias, law, validation, and feedback.
3. You receive a structured report detailing all deficiencies found—for instance, 'CRITERIA_ABSENT' or 'LEGAL_NONCOMPLIANCE'—and must revise the underlying logic to pass.

## Frequently Asked Questions

**Does the People Management Prover handle international laws?**
Yes. The tool verifies compliance against multiple jurisdictions, including US Title VII, GDPR Art. 22 for the EU, and CLT for Brazil, ensuring your policy respects local law.

**How does People Management Prover fix 'culture fit' bias?**
It rejects 'culture fit.' The tool forces you to replace that concept with defined behavioral indicators (KSAs) that trace back to the essential job functions, making it measurable.

**Is People Management Prover only for hiring?**
No. While focused on hiring and assessment, it also validates performance reviews and disciplinary actions, ensuring feedback is developmental and legally sound.

**What happens if I ignore the warnings from validate_people_management?**
The tool will reject the output and require you to fix structural deficiencies. It won't let you conclude a decision until criteria are defined, bias is audited, and legal checks pass.

**How do I input candidate data for the `validate_people_management` tool?**
You provide raw, mixed inputs: job descriptions, interview transcripts, and performance scores. The tool doesn't require pre-formatted files; it analyzes the text you feed it to identify gaps in your reasoning.

**Is my sensitive HR data safe when running `People Management Prover`?**
Yes, all data is processed through Vinkius' secure infrastructure. The connection uses standard encryption protocols, ensuring confidentiality for the records you pass to the agent.

**Can the `validate_people_management` tool process large batches of candidates efficiently?**
The tool is designed for structured assessment and can handle high volumes. However, keep in mind that complex analyses (like full adverse impact audits) require sufficient time to run fully.

**What specific output does the `validate_people_management` tool provide?**
The output is a structured critique detailing exactly which criteria are missing or flawed. It returns explicit codes like CRITERIA_ABSENT or FEEDBACK_EMPTY, telling you precisely where your reasoning fails.

**Why does the prover reject 'culture fit' as an evaluation criterion?**
Because culture fit is a subjective construct that correlates with interviewer similarity bias and lacks predictive validity, acting as a proxy for unlawful discrimination.

**How does it audit for adverse impact?**
By calculating selection rates for different groups and applying the EEOC 4/5ths rule across the recruitment funnel stages to highlight significant differences.

**What legal frameworks does this prover cover?**
It covers US federal anti-discrimination laws (Title VII, ADA, ADEA), EU regulations (GDPR Art. 22 and EU AI Act), UK Equality Act 2010, and Brazilian labor laws (CLT, Lei 9.029, LGPD).