CHRO / Chief Human Resources Officer

ReviewAudited by ClawScan on May 1, 2026.

Overview

This instruction-only HR advisor is coherent and disclosed, but users should handle any employee records or HR analytics it processes with care.

This looks safe to install as an instruction-only HR guidance skill. Before using it with real HR data, limit the records you share, avoid broad HRIS/personnel access unless necessary, and require human HR/legal review for employment, compensation, and compliance decisions.

Findings (2)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

If used with real employee data, the agent may process salaries, manager notes, activity patterns, and engagement information that could affect how employees are evaluated.

Why it was flagged

The skill expects sensitive employee records and inferred signals to be summarized into risk scores. This is aligned with HR analytics, but it involves private personnel data and outputs that should not be over-trusted or broadly reused.

Skill content
Attrition Risk Scoring ... 1:1 sentiment (from manager notes) ... Activity patterns ... Compensation vs market ... Output: Risk score
Recommendation

Provide only the minimum necessary HR data, anonymize where possible, restrict access to need-to-know users, and have HR/legal review sensitive conclusions before action.

What this means

Incorrect or incomplete HR data could lead to unnecessary escalation, delayed action, or unfair attention toward an employee.

Why it was flagged

The skill encourages HR alerts that could influence urgent employment workflows. This is disclosed and purpose-aligned, but inaccurate inputs or biased scoring could propagate into HR decisions if treated as authoritative.

Skill content
Proactive Alerts ... Urgent (same day): ... Missing documentation for pending termination ... High priority ... Employee shows high flight risk score
Recommendation

Treat alerts as prompts for human review, verify source data before acting, and keep the listed human-approval steps for terminations, compensation exceptions, and restructures.