internal-comms

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [Prompt Injection] (SAFE): The instructions are strictly focused on formatting internal communications like newsletters and FAQs. There are no attempts to override system prompts or bypass safety filters.
  • [Data Exposure & Exfiltration] (SAFE): No sensitive file paths, hardcoded credentials, or network-bound operations were detected. The skill references local documentation files in an examples/ directory.
  • [Remote Code Execution] (SAFE): The skill does not contain any scripts, package manifests, or commands to download/execute remote content.
  • [Indirect Prompt Injection] (SAFE):
  • Ingestion points: User-provided text intended for internal communication drafts.
  • Boundary markers: None explicitly defined in the top-level SKILL.md.
  • Capability inventory: None. The skill does not possess the ability to write files, execute commands, or access the network.
  • Sanitization: Not applicable as the skill has no functional side-effects beyond text generation.
  • [Obfuscation] (SAFE): No encoded strings, hidden characters, or homoglyphs were found in the text or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:28 PM