The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch the latest information from external Cloudflare documentation and local configuration schemas to guide command generation and execution. This creates an attack surface where compromised external content could influence agent behavior.
Ingestion points: External URLs https://developers.cloudflare.com/workers/wrangler/ and https://developers.cloudflare.com/workers/, and local file node_modules/wrangler/config-schema.json (SKILL.md).
Boundary markers: None provided in the instructions to the agent.
Capability inventory: Execution of shell commands via wrangler, npm, and npx; network requests via curl (SKILL.md).
Sanitization: No explicit sanitization or validation of the retrieved content is mentioned.
[EXTERNAL_DOWNLOADS]: The skill instructs the installation of the wrangler CLI and associated development tools from the official npm registry. These originate from Cloudflare, a well-known service.
[COMMAND_EXECUTION]: The skill provides numerous examples and instructions for executing shell commands, including project initialization, deployment, database management, and secret handling. It includes proactive warnings against unsafe secret management practices.

wrangler