The Agent Skills Directory

[DYNAMIC_EXECUTION]: The __init__.py file and the test suite (tests/test_secure_fetcher.py) utilize importlib.util and exec_module to dynamically load secure-fetcher.py. This is done to handle a script filename containing a hyphen, which prevents standard Python import syntax.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to fetch data from external URLs provided by the user or agent, which may contain malicious instructions targeting the LLM. * Ingestion points: External HTTP/HTTPS response data fetched via urllib3 in secure-fetcher.py. * Boundary markers: None identified; the skill returns the raw response body as a string. * Capability inventory: Network read operations via urllib3.PoolManager.request. * Sanitization: The content is decoded as UTF-8 with errors='replace', but no structural validation or sanitization for LLM instructions is performed.

secure-url-fetcher