The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/ddg_search.py uses subprocess.call to re-execute itself within a local virtual environment if it is not already active. This is a standard mechanism for ensuring dependencies are available at runtime.
[EXTERNAL_DOWNLOADS]: The scripts/bootstrap_venv.sh script installs the ddgs package from the Python Package Index (PyPI) to handle search queries. This is an expected dependency for the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The tool retrieves search results, news, and instant answers from DuckDuckGo. These results are external, untrusted data that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Search results and snippet data retrieved in scripts/ddg_search.py via the ddgs library and DuckDuckGo API.
Boundary markers: The skill returns data in structured JSON format by default, which helps the agent distinguish between query metadata and content, though it lacks explicit warnings to ignore instructions found within the search content.
Capability inventory: The skill is authorized to run the ddg-search script via the Bash tool.
Sanitization: The script does not filter or sanitize the text content of search results before returning them.

ddg-search