web-search
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
scripts/setup.shscript downloads the officialsearxng/searxngimage from Docker Hub to provide the search backend.\n- [COMMAND_EXECUTION]: The skill executes shell commands to manage Docker containers (docker run,docker start) and creates a local configuration directory at~/.config/searxng.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes snippets and content from third-party websites.\n - Ingestion points: The
SKILL.mdfile instructs the agent to fetch search results from the local SearXNG API instance.\n - Boundary markers: Absent. The instructions do not define delimiters or provide warnings for handling the embedded search result data.\n
- Capability inventory: The agent is empowered to use
web_fetchorbrowsertools to interact with web content based on search results.\n - Sanitization: Absent. There is no evidence of filtering, escaping, or sanitizing the retrieved search content before it is processed by the agent.
Audit Metadata