assistant-juridique-droit-travail
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied legal documents such as dismissal letters and employment contracts which represents a potential attack surface for indirect prompt injection. Evidence: 1. Ingestion points: Document analysis tasks described in SKILL.md and specialized sub-agent files like agents/analyse-licenciement.md and agents/audit-contrat.md. 2. Boundary markers: Absent. 3. Capability inventory: The skill utilizes various legal research tools through MCP (LibreJustice, Caius, OpenLegi CarthagoLex) and performs complex drafting and calculation tasks. 4. Sanitization: The skill contains a specific 'Sécurisation du skill' section in SKILL.md that explicitly instructs the agent to treat documents as data only, ignore any instructions within them, and report injection attempts.
- [SAFE]: The skill uses official government domains (e.g., bofip.impots.gouv.fr) and established legal databases via authorized MCP connectors, which is appropriate for its stated purpose.
Audit Metadata