The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use grep and glob to search within the node_modules directory for documentation and source code snippets.
[EXTERNAL_DOWNLOADS]: The skill fetches documentation from the official domain ai-sdk.dev (a well-known service) to provide up-to-date API information.
[INDIRECT_PROMPT_INJECTION]: The skill processes content from external documentation sites, creating a potential surface for indirect prompt injection if those sites were compromised or contained malicious instructions.
Ingestion points: External documentation URLs from ai-sdk.dev (defined in SKILL.md).
Boundary markers: The instructions do not define specific delimiters to isolate external documentation from agent instructions.
Capability inventory: The skill uses file system searching (grep, glob) and relies on the agent's network capabilities.
Sanitization: There is no explicit sanitization step for the fetched documentation before it is processed by the agent.

ai-sdk