fetching-blocked-urls
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
curlcommands in the shell to interact with a remote service. It includes a shell script retry loop that usesfor,grep, andsleepto manage connection attempts. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external websites, creating a surface for indirect prompt injection.
- Ingestion points: Markdown content retrieved from
https://r.jina.ai/based on user-provided or agent-selected URLs. - Boundary markers: None; the skill does not define specific delimiters or instructions for the agent to ignore potential commands embedded in the fetched markdown.
- Capability inventory: Shell command execution via
curland subsequent processing of external data. - Sanitization: No sanitization or content filtering is performed on the retrieved data before it is presented to the agent.
- [EXTERNAL_DOWNLOADS]: Fetches web content through Jina AI's reader service (
r.jina.ai), which is a well-known service for converting web pages into clean markdown.
Audit Metadata