The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local configuration files and CLI outputs, which could theoretically contain malicious instructions if a project file is compromised.
Ingestion points: Processes content from wrangler.jsonc, cloudflare.config.ts, package.json, and the output of various cf commands.
Boundary markers: The instructions do not specify the use of delimiters or clear separation between ingested data and agent instructions.
Capability inventory: The skill utilizes cf CLI commands for infrastructure management (deploy, create, migrate), curl for network verification, and git diff for inspecting configuration changes.
Sanitization: There is no evidence of explicit sanitization, though the skill mandates anonymizing search queries to prevent data leakage.

cf-cli