The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and installs the @leeguoo/yapi-mcp package from the NPM registry and the leeguooooo/cross-request-master skill. It also triggers the installation of agent-browser-stealth to support browser-based authentication flows.
[COMMAND_EXECUTION]: The skill's primary functionality is implemented through shell commands, including the yapi CLI, npm/npx for installation and execution, and rg (ripgrep) for inspecting local configuration files.
[CREDENTIALS_UNSAFE]: The skill reads and manages sensitive YApi authentication data stored in the user's home directory at ~/.yapi-mcp/auth-*.json and configuration at ~/.yapi/config.toml.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it fetches and summarizes interface documentation from external YApi servers which may contain untrusted data.
Ingestion points: Untrusted data enters the agent context via yapi interface get, yapi search, and yapi interface list-menu in SKILL.md.
Boundary markers: The instructions lack delimiters or explicit instructions to ignore potentially malicious content embedded in the documentation.
Capability inventory: The skill allows shell command execution through the yapi CLI in SKILL.md.
Sanitization: No validation or sanitization is performed on the documentation content fetched from the remote server before the agent processes it.

yapi