copy
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data during the first-run setup interview and subsequently processes it to generate copy. Specific details such as business information, audience pain points, and testimonials are stored in the
./partners/directory and re-read during copy generation tasks. - Ingestion points:
SKILL.md(Interview and./partners/files). - Boundary markers: None present; the skill interpolates the stored profile data directly into its reasoning process without delimiters or instructions to disregard embedded commands.
- Capability inventory: The skill possesses file write and read capabilities restricted to the
./partners/directory. - Sanitization: No validation or sanitization of the input provided during the interview is performed.
- [EXTERNAL_DOWNLOADS]: The
README.mdfile contains instructions for users to download the skill's source code from a GitHub repository (https://github.com/tenfoldmarc/copy-skill). This is the official distribution channel for the skill authored by tenfoldmarc.
Audit Metadata