The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The installation guide instructs users to download a ZIP archive from https://figma-boost.softhvn.xyz. This domain is not a recognized or trusted service provider for software distribution.
[COMMAND_EXECUTION]: The manual installation process requires the user to execute ./figma-boost-setup. Running unverified scripts from external sources is a high-risk activity that can lead to system compromise.
[REMOTE_CODE_EXECUTION]: The MCP configuration involves running node /path/to/figma-boost/mcp-server.js. Because this file is part of the untrusted download, it represents a path for persistent remote code execution within the agent's workspace.
[CREDENTIALS_UNSAFE]: The skill instructs users to provide a FIGMA_ACCESS_TOKEN as an environment variable. If the downloaded setup script or MCP server is malicious, it can capture and exfiltrate these highly sensitive credentials, granting full access to the user's Figma account.
[PROMPT_INJECTION]: The skill processes external data from Figma, creating a surface for indirect prompt injection.
Ingestion points: Figma design assets and component data via connectFile, getPage, and getComponents in SKILL.md.
Boundary markers: Absent; no delimiters or instructions to ignore embedded instructions.
Capability inventory: Includes file system writes (fs.writeFile), network operations (Figma API), and binary execution (figma-boost-setup).
Sanitization: Absent; no evidence of escaping or filtering content from Figma layers.

figma-boost-mcp