The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content retrieved from external websites and search results, which is then fed into the agent's context. This creates a surface for indirect prompt injection attacks.
Ingestion points: Content from DuckDuckGo search results and arbitrary URLs via WebFetch or curl.
Boundary markers: No specific boundary markers or instructions are provided to the agent to treat external content as untrusted.
Capability inventory: The skill is configured to allow the Bash tool, providing a potential execution path if the agent follows malicious instructions from the web.
Sanitization: There is no evidence of sanitization or filtering of the external content.
[EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch search data from DuckDuckGo and SerpAPI, which are recognized well-known services.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute curl, jq, and grep for the purpose of searching the web and parsing data.

web-search