The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: References official Litestar framework documentation from well-known domains to provide technical context.\n- [CREDENTIALS_UNSAFE]: Uses development-only placeholder strings like 'dev-token' and 'replace-in-production' in documentation examples, with explicit instructions to replace them in production environments.\n- [INDIRECT_PROMPT_INJECTION]: Documents patterns for ingesting data from HTTP headers and path parameters in 'references/security-patterns.md'. Boundary markers are established via separate request parsing and auth context layers. Capabilities focus on identity and authorization logic. Sanitization is addressed through prefix removal and deterministic, side-effect-free guards.\n- [SAFE]: Promotes strong security hygiene by advocating for constant-time secret comparisons and the use of dedicated data structures for handling sensitive information.

litestar-security