The Agent Skills Directory

[SAFE]: The skill provides legitimate documentation, implementation templates, and linter-like rules for Cloudflare R2 storage. It correctly identifies security risks such as client-side credential exposure and mitigates them by recommending server-side signing and secret management via environment variables. The skill also incorporates neutral documentation for Cloudflare as a well-known service, including guides for handling platform-specific errors and CORS configurations.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection attacks by facilitating the ingestion of untrusted data through R2 storage operations.
Ingestion points: The skill ingests file content and metadata via put, get, and multipart upload methods, as seen in templates/r2-simple-upload.ts and templates/r2-multipart-upload.ts.
Boundary markers: The provided code templates do not implement delimiters or specific instructions to the agent to disregard potential commands embedded within the stored data.
Capability inventory: The skill grants the agent capabilities to perform file-read, file-write, and file-delete operations within the R2 environment.
Sanitization: While some examples (e.g., in references/common-patterns.md) include basic file type validation, the skill does not enforce sanitization of file content to prevent instruction injection.

cloudflare-r2