The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources that could contain malicious instructions designed to influence the agent's behavior in subsequent tasks.
Ingestion points: The skill fetches data via MCP tools from lexsis_catalog.review_collections and performs "web search or browser research" for public facts.
Boundary markers: The instructions lack explicit delimiters (e.g., XML tags or triple quotes) to wrap ingested data, although it requires marking inferences and recording source URLs.
Capability inventory: The agent has the ability to write files to the work/storefront/ directory and execute various MCP tool calls for storefront management.
Sanitization: There is no requirement to sanitize or filter the content of customer reviews or web search results before writing them to context files.
[EXTERNAL_DOWNLOADS]: The skill connects to an external MCP server at https://mcp.trylexsis.com/mcp to retrieve storefront data. This connection is fundamental to the skill's stated purpose of integrating with the Lexsis platform.

setup