parallel-web
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted web content from searches and URL extractions, which may contain malicious instructions designed to manipulate the agent.
- Ingestion points: Synthesized search summaries, citation excerpts, and full web page content retrieved via the Parallel APIs in
scripts/parallel_web.py. - Boundary markers: The skill does not implement explicit delimiters or instructions in its prompts to ensure the agent treats retrieved web data as potentially untrusted.
- Capability inventory: The skill environment is configured with
Read,Write,Edit, andBashtools, which could be exploited if an injection attack is successful. - Sanitization: The scripts do not perform specific sanitization or filtering of the retrieved web data before passing it to the agent.
Audit Metadata