skills/mohn93/ff-mcp/community-ff-mcp/Gen Agent Trust Hub

community-ff-mcp

Warn

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct users to install the 'community-ff-mcp' package using npx. This results in the download of code from the npm registry at runtime.
  • [REMOTE_CODE_EXECUTION]: The prerequisite setup requires running 'npx -y community-ff-mcp', which executes remote code on the user's host machine.
  • [COMMAND_EXECUTION]: The skill provides shell commands for adding the MCP server to the AI client environment using 'claude mcp add'.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: Untrusted data enters the agent context through tools like 'get_page_by_name', 'get_custom_code', and 'get_project_yaml' which fetch YAML and code from the external FlutterFlow API (SKILL.md).
  • Boundary markers: There are no instructions or delimiters defined to prevent the agent from following instructions potentially embedded within the retrieved YAML or custom code.
  • Capability inventory: The skill possesses capabilities to write to the local file system (sync_project) and push project modifications back to the external API (update_project_yaml).
  • Sanitization: There is no evidence of validation, filtering, or sanitization of the content fetched from the API before it is processed by the AI assistant.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 14, 2026, 05:44 PM
Security Audit — agent-trust-hub — community-ff-mcp