The Agent Skills Directory

[SAFE]: The skill is a well-structured set of documentation and code examples for the Cloudflare Workers platform. It follows security best practices by explicitly mandating the use of parameterized SQL queries to prevent SQL injection in D1 databases, using streaming for R2 object storage to prevent memory exhaustion, and utilizing official Cloudflare bindings for service interactions. It provides clear guidance on managing secrets using wrangler secret and environment-specific configurations. No malicious patterns, obfuscation, or unauthorized data exfiltration techniques were detected.
[INDIRECT_PROMPT_INJECTION]: The skill provides architecture for building edge applications that process untrusted data from HTTP requests. Ingestion points: Requests are parsed via request.json(), request.body(), and URL parameters across various handler examples. Boundary markers: The instructions mandate typed handlers (satisfies ExportedHandler<Env>) and demonstrate structured message schemas for AI interactions. Capability inventory: The skill provides patterns for persistent storage (D1, KV, R2), background processing (Queues), and AI inference (Workers AI). Sanitization: The skill enforces rigorous security controls, including parameterized SQL queries (prepare().bind()), input validation (length and content-type checks), and streaming to prevent memory-based denial of service.

infra-platform-cloudflare-workers