ai-companion

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
  • [NO_CODE]: The skill consists solely of instructions and configuration files. It contains no Python, JavaScript, or shell scripts.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes external data from a user-modifiable file to determine its logic.
  • Ingestion points: Processes metadata from ~/.agents/skills/skills.json (SKILL.md).
  • Boundary markers: Absent; there are no instructions to the agent to disregard potential commands or overrides embedded within the JSON data.
  • Capability inventory: The skill has the ability to influence the agent's selection and execution of other skills (control flow manipulation).
  • Sanitization: Absent; the skill does not perform validation or sanitization on the contents of the taxonomy file.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 09:17 PM
Security Audit — agent-trust-hub — ai-companion