arabic-web-search
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
curlto interact with search engine APIs (DuckDuckGo and SearXNG). These commands are executed by the agent to retrieve external data based on user queries. - [EXTERNAL_DOWNLOADS]: Data is fetched from external domains including
api.duckduckgo.comandsearx.beto provide search results. - [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection (Category 8).
- Ingestion points: Untrusted search results from the internet are ingested into the agent context via
curloutput. - Boundary markers: Absent. The instructions do not include specific delimiters or warnings for the agent to ignore instructions that might be present in the search snippets.
- Capability inventory: The agent has the capability to execute shell commands (
curl) as defined in the skill. - Sanitization: No sanitization or filtering of the search results is performed before the agent processes them.
Audit Metadata