The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and analyze user-provided Django source code and project architecture for review and refactoring purposes. This ingestion of external data constitutes a surface for indirect prompt injection.
Ingestion points: User-provided Django features, models, views, and settings ingested via SKILL.md and references/best-practices.md.
Boundary markers: The instructions do not currently define explicit delimiters or instructions for the agent to ignore natural language within code blocks.
Capability inventory: The skill references shell command execution capabilities via standard python manage.py tools listed in references/workflows.md.
Sanitization: No explicit sanitization or input validation is specified for user-provided code snippets.
Mitigation: To harden the skill, external code content should be wrapped in unique delimiters with instructions for the agent to treat the content as data only.
[COMMAND_EXECUTION]: The skill documents standard Django administrative commands in references/workflows.md, including runserver, makemigrations, migrate, test, and collectstatic. These commands are integral to the framework's operation and are presented in a legitimate development context without suspicious flags or obfuscation.

django