The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local shell script scripts/crawlora.sh to execute curl commands for API interaction. This is standard functionality for communicating with the vendor's REST API and handles environment-variable-based authentication safely.- [EXTERNAL_DOWNLOADS]: All network operations are directed to api.crawlora.net, which is the official API endpoint for the skill's author. These downloads are restricted to Lululemon catalog and store data.- [INDIRECT_PROMPT_INJECTION]: The skill retrieves untrusted third-party content in the form of product reviews through the lululemon_product endpoint, representing a potential surface for indirect prompt injection.\n
Ingestion points: The product_id detail endpoint in reference/endpoints.md.\n
Boundary markers: The skill instructions do not specify the use of delimiters or markers to isolate fetched review content within the agent's prompt context.\n
Capability inventory: Shell command execution via curl in scripts/crawlora.sh.\n
Sanitization: No explicit sanitization or filtering of the review text is implemented within the skill's scripts.

lululemon-research