barry-logic-to-speech

Warn

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill directs the agent to search for and read files from several sensitive application configuration directories, including ~/.claude/skills/, ~/.cursor/skills/, ~/.codex/skills/, and ~/.agents/skills/, which could expose information about the user's environment.
  • [DATA_EXFILTRATION]: The skill performs persistent writes to the local file system. It saves extracted vocabulary terms to ~/.barry-english/fancy-vocab.md and records full English speech transcripts to ~/.barry-english/logic-to-speech-output-YYYY-MM-DD.md.
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface (Category 8) because it processes untrusted user data without sufficient safeguards.
  • Ingestion points: User-provided raw information, logic prompts, and speech materials processed in SKILL.md.
  • Boundary markers: Absent; the skill does not define clear delimiters or "ignore instructions" warnings for user-provided data.
  • Capability inventory: File system read access (multiple application-specific directories) and file system write access (local vendor directory).
  • Sanitization: Absent; the skill does not implement validation or escaping for the external content before it is processed or saved to disk.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 14, 2026, 02:44 PM
Security Audit — agent-trust-hub — barry-logic-to-speech