generate-import-html

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests untrusted data from external webpages.
  • Ingestion points: cleaned.html and metadata.json (scraped from external sources in previous steps).
  • Boundary markers: None identified. There are no instructions for the agent to distinguish between its own system instructions and potentially malicious instructions embedded in the ingested content.
  • Capability inventory: The skill has the capability to write files (paths.htmlFilePath), create directories (mkdir), and copy files (cp).
  • Sanitization: None. The skill is explicitly instructed to include all text and structure from the source files without omission or modification.
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform file system operations (mkdir -p, cp -r) using paths derived from variables in the metadata.json file. Specifically, paths.htmlFilePath is used to determine the directory structure for saving HTML and images. Without explicit validation or sanitization of these paths in the instructions, a maliciously crafted metadata file could be used to facilitate path traversal attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 04:35 AM