The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches and processes content from the web via DuckDuckGo, representing a surface for indirect prompt injection attacks where untrusted data could influence agent behavior.
Ingestion points: The scripts/search.py script retrieves search results including titles and descriptions from an external API.
Boundary markers: The output results are formatted for readability but lack explicit delimiters or instructions to ignore embedded commands.
Capability inventory: The skill can make network requests and write search results to local files via the --output command-line argument.
Sanitization: There is no evidence of sanitization or filtering of the search result content before it is presented to the agent.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of a third-party Python package to function.
Evidence: SKILL.md instructs users to install duckduckgo-search using pip.

web-search