The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation relies on the npx ai-cloudflare command, which fetches the ai-cloudflare package from the public npm registry at runtime.
[REMOTE_CODE_EXECUTION]: Execution via npx constitutes remote code execution as the package is downloaded and executed in the local environment. The package ai-cloudflare is provided by 'LXGIC Studios' (referenced as the authoring studio in the text), which is not listed as a trusted vendor or well-known service.
[COMMAND_EXECUTION]: The skill's primary functionality is delivered through shell command execution, specifically using the npx binary to run external tools.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it transforms untrusted user descriptions into executable code and configuration.
Ingestion points: User-provided plain English descriptions in the npx command arguments (e.g., in SKILL.md).
Boundary markers: No boundary markers or 'ignore' instructions are used to delimit user input from the generation logic.
Capability inventory: The tool generates Cloudflare Worker code capable of network operations, KV storage access, and R2 bucket interactions.
Sanitization: There is no evidence of input sanitization or validation of the generated code before it is presented to the user or deployed.

cloudflare-gen