The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions involve running standard development commands such as Django management commands, linters (ruff, black, isort), and test runners (pytest). These are necessary for the skill's stated purpose of project verification.
[CREDENTIALS_UNSAFE]: The skill includes steps to verify the existence of sensitive environment variables like DJANGO_SECRET_KEY and utilizes gitleaks to proactively search for hardcoded secrets. It also provides a template command for superuser creation which uses placeholder credentials.
[PROMPT_INJECTION]: The skill processes local source code and git diffs, representing an indirect prompt injection surface. However, this risk is mitigated by the skill's specific context as a developer utility.
Ingestion points: Project source code, configuration files, and git diff output.
Boundary markers: None.
Capability inventory: Shell access for execution of linters, security scanners, and test suites.
Sanitization: Not applicable, as the skill operates on the project's own code for auditing purposes.

django-verification