The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The configuration instructions in references/configs.md guide users to run the MCP server using npx -y zephex. This command downloads and executes the zephex package from the npm registry at runtime. This is the intended method for running the vendor's tool.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the zephex:inspect_url tool.
Ingestion points: The tool zephex:inspect_url fetches content from any public URL provided by the user or found in the codebase (documented in references/tools.md).
Boundary markers: No specific instructions are provided to delimit this external content or warn the agent to ignore instructions embedded within the fetched data.
Capability inventory: The agent has access to sensitive capabilities including reading local source code (zephex:read_code), performing repository-wide searches (zephex:find_code), and it is instructed to perform npm install operations after checking packages.
Sanitization: While the tool is described as 'cleaning' content, no specific security sanitization or validation of the fetched content is described to prevent prompt injection.

zephex