The Agent Skills Directory

[SAFE]: The skill demonstrates a strong security posture by providing explicit best practices for preventing common vulnerabilities. For example, it includes instructions and code examples for validating user-supplied URLs using Django's url_has_allowed_host_and_scheme to prevent open redirect attacks.
[SAFE]: It emphasizes the importance of rate limiting (throttling) on authentication endpoints, specifically advising users to run throttles before authentication logic to mitigate brute-force attempts.
[SAFE]: All external documentation and repository links point to official vendor domains and reputable services like readthedocs.io and github.com associated with the library's author, wemake-services.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch documentation from external URLs at runtime, which is an inherent surface for indirect prompt injection. However, given the primary purpose of the skill is library guidance and the sources are official project docs, the risk is minimal.
Ingestion points: Documentation is fetched from django-modern-rest.readthedocs.io in llms.txt or .md formats.
Boundary markers: No explicit delimiters are used for the remote content.
Capability inventory: The skill primarily guides code generation and does not execute dangerous local subprocesses.
Sanitization: No specific sanitization of the remote documentation content is described.

dmr