file-to-markdown
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- External Download and Setup: The skill includes commands to fetch a Java Runtime Environment (JRE) from the official Adoptium (Eclipse Foundation) API if Java is not already present on the system. The instructions specifically advise the agent to obtain user permission before performing any installation, which is a positive security practice for managing external dependencies.
- Indirect Prompt Injection Surface: As the skill's primary function is to ingest and process content from a wide variety of untrusted file formats (such as PDF, Word, and Excel), it is susceptible to indirect prompt injection if the processed files contain malicious instructions. The skill mitigates this risk by recommending structured JSON output (
rmeta) and suggesting that output be redirected to files rather than being directly injected into the agent's primary context, which helps maintain operational discipline.
Audit Metadata