docx

Pass

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists of documentation and best practices for processing Word documents. It does not include any executable scripts, remote downloads, or hidden commands.
  • [NO_CODE]: No implementation scripts or executables are provided within the skill package. All instructions are purely textual guidelines for the agent.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructions involve reading and processing external .docx files, which are a source of untrusted data.
  • Ingestion points: Document reading and content manipulation steps described in SKILL.md.
  • Boundary markers: None provided in the instructional text.
  • Capability inventory: The skill is configured to use the 'exec' tool for shell operations as defined in _meta.json.
  • Sanitization: No specific sanitization or validation steps for external content are included in the guidelines.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 15, 2026, 03:23 AM
Security Audit — agent-trust-hub — docx