The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The generate_schema.py utility ingests schema definitions in JSON format to bootstrap Python source code. This creates a vulnerability where maliciously crafted schema names or field properties could be used to inject arbitrary logic into the generated code.
Ingestion points: The --entities command-line argument in scripts/generate_schema.py processes external data strings.
Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within the ingested JSON data.
Capability inventory: The skill possesses capabilities for directory creation and file writing on the local file system.
Sanitization: The script performs standard JSON parsing but lacks validation to ensure entity names and field types match safe Python identifier patterns before interpolation into code templates.
[DYNAMIC_EXECUTION]: The skill implements a template-based generation system that constructs Python source files at runtime.
Evidence: The script uses string concatenation and f-strings to build executable Python files for ORM models and API routes.

database-schema