clean-content-fetch

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructions and command examples hardcode absolute file paths targeting a specific user directory (/Users/zzd/). This exposes the local system's directory structure and the username to the agent.
  • [COMMAND_EXECUTION]: The skill operates by executing a local Python script through the shell, passing user-supplied input (the target URL) directly as a command-line argument.
  • [EXTERNAL_DOWNLOADS]: The documentation instructs the agent to install several external Python packages (scrapling, html2text, curl_cffi, playwright, browserforge) and to download browser binaries using the Playwright CLI.
  • [PROMPT_INJECTION]: The skill functions as a gateway for untrusted data by fetching content from arbitrary external URLs (Indirect Prompt Injection surface). Malicious instructions embedded in a fetched web page could potentially influence the agent's subsequent actions.
  • Ingestion points: External content enters the context via the scrapling_fetch.py script output.
  • Boundary markers: The skill does not define specific delimiters or instructions to the agent to ignore embedded commands within the fetched content.
  • Capability inventory: The agent has the ability to execute shell commands and perform network requests.
  • Sanitization: There is no evidence of sanitization or filtering of the fetched content to prevent instruction injection before it is returned to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 03:36 AM