Purpose
Turn a messy set of case facts into a clean, standardized employee relations case record — correctly categorized, risk-rated, and stamped with a clear next action and owner — so cases don't stall, escalation criteria are consistent, and any HRBP or legal partner can pick up the file cold.
When to Use
Invoke this skill when the user asks to:
- Open or log a new employee relations case
- Triage an incoming complaint, grievance, or concern
- Document an investigation, disciplinary matter, or accommodation request
- Update the status, risk level, or next action on an existing case
- Summarize a case for a manager, HR leader, or legal handoff
- Assess whether a case needs escalation or outside counsel
Inputs Required
Before starting, collect:
- Case description: 3–6 sentences describing what happened, in the user's words
- Case type (if known): e.g., harassment, discrimination, grievance, performance/conduct, accommodation, wage/hour, retaliation, policy violation, workplace safety
- Jurisdiction: Country, and state/province if U.S./Canada (drives legal exposure)
- Parties involved: Complainant, respondent, witnesses, and their roles/levels
- Open date: When the case was reported or opened (default to today if unstated)
- Current status: New / Under investigation / Awaiting decision / Remediation / Closed
- Any actions already taken: Interviews done, interim measures, notifications made
If case type or risk level isn't provided, infer it from the description and label it as an inference for the user to confirm.
Instructions
Step 1: Categorize the Case
Assign a primary case type and, if applicable, a secondary type (many cases are cross-cutting — a "grievance" may also carry a retaliation component).
Primary categories:
- Harassment/Hostile Environment — unwelcome conduct tied to a protected characteristic
- Discrimination — adverse treatment based on a protected class
- Retaliation — adverse action following protected activity (complaint, leave, whistleblowing)
- Conduct/Behavioral — insubordination, dishonesty, violence, substance issues
- Performance — capability and results issues (distinct from conduct)
- Grievance/Working Conditions — dispute over treatment, workload, management, or policy application
- Accommodation — disability, religious, or medical accommodation request
- Wage/Hour — pay, overtime, classification, or leave-pay disputes
- Policy Violation — code of conduct, attendance, IT/security, or expense breaches
- Safety/Whistleblower — hazard reports or protected disclosures
Tag whether the case involves a protected characteristic or protected activity — this is the single biggest driver of legal risk.
Step 2: Assess Risk Level
Rate the case Low / Medium / High / Critical using these signals:
- Critical: Alleged illegality (assault, fraud, safety threat to life), executive or officer respondent, active legal claim or agency charge (EEOC, tribunal, DOL), media/reputational exposure, or credible threat of violence
- High: Involves a protected characteristic or protected activity, potential termination, multiple complainants, pattern of prior complaints against the respondent, or accommodation with documented medical basis
- Medium: Interpersonal conflict, first-time policy violation, performance dispute with no protected-class angle, or grievance about a manageable working condition
- Low: Minor policy clarification, coaching-level concern, or resolved-on-contact issue
State the one or two facts driving the rating. If any Critical or High signal is present, set an Escalation Flag: YES and name who it should go to (senior HR leadership, in-house or outside counsel, safety/security).
Step 3: Map Jurisdiction Considerations
Note jurisdiction-specific factors that affect handling — without giving legal advice:
- U.S.: At-will status, applicable state protections beyond federal (e.g., CA, NY, IL have broader coverage), agency filing windows
- UK/EU: Statutory grievance/disciplinary procedures, notice and consultation requirements, works councils
- Canada: Provincial employment standards, just-cause thresholds, human rights codes
- Other/Multi-site: Flag if the case spans jurisdictions and note that local requirements govern
Add a one-line note: "Confirm local legal requirements before any disciplinary or separation decision." when risk is Medium or above.
Step 4: Build the Timeline
Create a dated event log from the facts provided:
- Open date and how the case was reported
- Any interim measures taken (separation of parties, suspension, access changes)
- Interviews conducted (party + date)
- Decisions or notifications made
If dates are missing, list events in sequence and mark undated items as "date TBD."
Step 5: Set the Next Action
Define one clear next action, its owner, and a target date. The next action should move the case toward resolution — not restate the status. Examples: "Interview respondent (owner: J. Rivera, by [3 business days from open])" or "Send accommodation interactive-process letter (owner: HRBP, by Friday)."
Output Format
- Structure: A single structured case record with a header block, a details table, a timeline, and a summary paragraph
- Length: 350–600 words
- Format: Field table for case data; bulleted timeline; short prose summary
- Sections: 1. Case Record Header — Case ID (generate as ER-<year>-<3-digit sequence, use 001 if unknown>), Case Type (primary/secondary), Status, Open Date, Risk Level, Escalation Flag 2. Case Details Table — Jurisdiction, Parties (with roles), Protected characteristic/activity flag, Interim measures 3. Timeline — dated event log 4. Case Summary — one neutral, factual paragraph a leader or lawyer could read cold 5. Next Action — action / owner / target date 6. Escalation Note — who to notify and why (only if flag is YES)
- Tone: Neutral, factual, and non-conclusory — document allegations as "alleged," avoid characterizing guilt or credibility
Quality Checklist
Before delivering output, verify:
- [ ] A primary case type is assigned (and secondary if the facts warrant)
- [ ] Risk level is rated with the specific fact(s) driving the rating stated
- [ ] Escalation Flag is set to YES or NO — never left blank
- [ ] The Next Action names a specific action, an owner, and a target date
- [ ] Allegations are written as allegations ("alleged," "reported") — no conclusions of guilt
- [ ] Jurisdiction is captured and a legal-confirmation note appears when risk is Medium or above
Notes
- This skill documents and triages cases; it does not render legal conclusions or determine outcomes. Any statement about wrongdoing should remain framed as an allegation until an investigation concludes.
- The protected-characteristic/protected-activity flag is the highest-leverage field — when in doubt, flag it and rate risk higher, not lower.
- For updates to an existing case, keep the original Case ID and open date, append new timeline entries, and re-evaluate risk level and next action rather than overwriting them.
- Encourage the user to store the record in their case management system of record; this output is a structured entry, not a substitute for that system.
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills