assignation-refere-communication-associe-selim-brihi

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill facilitates the creation of legal documents based on the French Commercial Code. It provides guidance for information collection and document structure in a manner consistent with professional legal assistance.
  • [PROMPT_INJECTION]: No attempts to override agent instructions, bypass safety filters, or extract system prompts were found in the skill or its reference files.
  • [DATA_EXFILTRATION]: The skill does not perform any network operations, nor does it attempt to access sensitive local files such as credentials, SSH keys, or environment variables.
  • [INDIRECT_PROMPT_INJECTION]: The skill features a data ingestion surface through user input during the document drafting process.
  • Ingestion points: User responses to the 12-question workflow defined in references/workflow-informations.md.
  • Boundary markers: Absent; there are no specific delimiters to separate user data from instructions.
  • Capability inventory: The skill facilitates the generation of a .docx file (Phase 4 of SKILL.md). No network or shell execution capabilities are utilized.
  • Sanitization: No explicit sanitization of user-provided information is mentioned.
  • Note: This is a standard surface for a document-generation tool and is handled within the scope of its primary purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 04:12 AM
Security Audit — agent-trust-hub — assignation-refere-communication-associe-selim-brihi