From vanguard-frontier-agentic
Defines a structured legal-hr-case-capsule for auditable cross-agent handoffs with redacted identifiers, risk labels, privilege/privacy posture, and a decision-owner field.
How this skill is triggered — by the user, by Claude, or both
Slash command
/vanguard-frontier-agentic:legal-hr-case-capsuleThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
This skill defines the **legal-hr-case-capsule** — the single structured record
This skill defines the legal-hr-case-capsule — the single structured record that Legal and HR agents exchange when a matter crosses an agent boundary. The capsule exists so that no agent works in a silo, no context is lost in a handoff, and every cross-domain exchange is auditable. It is a data contract, not a decision: a capsule never approves, denies, or directs an action. It carries facts, uncertainty, risk posture, and an explicit decision owner.
decision_owner (an accountable human)
and exactly one primary_agent.do_not_do_list. An empty list is not acceptable;
if nothing is prohibited, the capsule is not ready to send.privilege_sensitivity and privacy_sensitivity.assumptions, inferences, and
missing_evidence are mandatory fields, not optional.risk_rating uses risk language only.escalation_required: true and
recommended_next_action to a pause-and-escalate posture unless documented
controls already exist.The capsule has 30 required fields. See references/capsule-schema.md for the full field definitions, allowed values, and redaction rules. Summary:
case_id, source_agent, receiving_agent,
primary_agent, secondary_agents, matter_type,
employee_or_party_identifiers_redacted, jurisdiction_or_location_if_known,
business_unit, timeline.facts, allegations, assumptions, inferences,
missing_evidence, evidence_quality.risk_rating, privilege_sensitivity, privacy_sensitivity,
retaliation_risk, discrimination_or_harassment_risk, regulatory_risk,
litigation_hold_needed, data_minimization_notes.decision_owner, human_approval_required,
escalation_required, recommended_next_action, do_not_do_list,
audit_log_summary.Load only when needed:
npx claudepluginhub raishin/vanguard-frontier-agentic --plugin vanguard-frontier-agenticCreates a structured, auditable handoff capsule when a Salesforce specialist agent transfers a matter to another agent, preserving context, risk labels, privilege posture, and privacy labels.
Interactive setup for employment-law jurisdiction mapping. Reads handbooks and termination memos, builds an escalation table, and configures CLAUDE.md. Use on fresh install or with --redo.
Walks through a Data Subject Access Request (access, deletion, portability, correction) and drafts the response — verify identity, locate data system-by-system, assess exemptions, draft acknowledgment and substantive letters.