The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to interact with the MCP server using curl within a shell environment (as seen in the Workflow section of SKILL.md). This creates a vulnerability where malicious user input could be used to perform command injection if the agent interpolates search keywords directly into the shell string without proper escaping.
[EXTERNAL_DOWNLOADS]: The skill's core functionality depends on fetching data from an external, third-party endpoint: https://yuju777-coupang-mcp.hf.space/mcp. This server is hosted on Hugging Face Spaces by an individual user and is not a verified platform-wide service.
[DATA_EXFILTRATION]: User-supplied search keywords are sent to the external server yuju777-coupang-mcp.hf.space. While functional for a search tool, this constitutes a data flow of user queries to a third-party infrastructure.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8).
Ingestion points: Data is received from the external MCP server at https://yuju777-coupang-mcp.hf.space/mcp (SKILL.md).
Boundary markers: None are defined to separate untrusted search results from the agent's instructions.
Capability inventory: The agent has the ability to execute shell commands (curl), perform network operations, and access the file system.
Sanitization: There are no instructions for sanitizing or validating the content returned by the external API before it is processed by the agent.

coupang-product-search