Back to Blog

automation

HR Assistant Skill: Automate Employee Queries in Agents

A scenario-driven guide to the hr-assistant skill for handling HR queries in AI agents, with installation and usage.

OpenAgentSkillPublished:

Where this fits

You've built an internal agent that employees use to ask about company policies. A new hire asks, "How many days of annual leave do I get after my probation?" Your agent needs to pull the exact answer from the employee handbook, confirm the employee's tenure, and maybe open a ticket if the employee wants to request leave. Manually coding each HR workflow is brittle. The hr-assistant skill gives your agent a predefined, step-by-step procedure to handle these queries consistently.

Why agents benefit

  • Structured knowledge retrieval: Uses kb_search with HR-specific keywords (e.g., "年假", "薪资") to find precise clauses from the employee handbook, so answers are grounded and auditable.
  • Employee context awareness: Calls customer_profile to fetch individual details like job level, hire date, or contract status—essential for personalized responses.
  • Actionable task handling: Creates HR tickets via create_ticket when the user needs a document, onboarding, or a process request, turning a Q&A into a workflow.
  • Policy freshness: Falls back to bocha_search for external labor law or social insurance updates, keeping answers current without manual maintenance.
  • Clear response rules: Forces a conclusion-first answer with citations ("依据员工手册第 X 条"), reducing ambiguity and improving user trust.

Practical scenarios

Employee asks about annual leave balance The agent confirms the user's identity via customer_profile, then queries kb_search for "年假" to get the accrual rules based on tenure. It replies with the exact entitlement and offers to open a leave request ticket.

New hire asks about onboarding paperwork The agent uses "入职" and "试用期" keywords to pull the onboarding checklist, then creates a ticket to pre-fill forms. It also checks the employee's probation end date from the profile to tailor the response.

Manager asks about salary adjustment policy The agent searches "调薪" in the handbook, cross-references the manager's own profile for eligibility, and if needed, queries external sources for recent compensation law changes before giving a definitive answer.

Add it to your agent workflow

Install the skill with:

npx skills add zj-unicom-ai/UniEmployee --skill hr-assistant

Then, in your agent's system prompt, reference the skill when HR topics appear. For example:

If the user asks about HR policies, employee benefits, or personnel processes, follow the hr-assistant skill steps: confirm intent, search the knowledge base, query employee profile if needed, and create a ticket when required.

Compare before adopting

Before integrating, evaluate the skill's quality signals: the GitHub repo has 21 stars—modest but active. Check the README for maintenance freshness (last commit date) and whether the knowledge base structure matches your company's handbook format. Compare with alternatives like building a custom RAG pipeline—this skill saves time if you need a quick, rule-based HR handler. Also assess workflow fit: does your agent already have kb_search and customer_profile functions? If not, you'll need to adapt the skill to your existing tools.

Why it is worth tracking

This skill addresses a common pain point—HR automation—with a clean, reproducible procedure. The low star count means it's early, but the author (zj-unicom-ai) demonstrates a practical approach. Track its updates to see if it adds more integrations or expands language support. Re-evaluate when your agent's HR needs grow or when the skill shows signs of active maintenance. For a deeper look, visit the skill page to monitor community feedback and usage patterns.

Featured Skill

hr-assistant

Skill by zj-unicom-ai