The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides patterns for ingesting external data through web scraping without demonstrating sanitization or the use of boundary markers. This represents a vulnerability surface for indirect prompt injection.
Ingestion points: modules/real-world-applications.md (fetches URL content using aiohttp)
Boundary markers: Not implemented in code examples
Capability inventory: Network requests via aiohttp, database operations, and task management
Sanitization: Not implemented in code examples
[DYNAMIC_EXECUTION]: The database interaction patterns in modules/real-world-applications.md use string interpolation (f-strings) to construct SQL queries. This is an unsafe practice that can lead to SQL injection vulnerabilities if the variables contain untrusted data.
[EXTERNAL_DOWNLOADS]: The skill references and suggests the installation of well-known Python packages including aiohttp and pytest-asyncio for its core functionality.

python-async