developing-genkit-dart

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install global packages via NPM (e.g., npm install -g genkit-cli) and add dependencies via Dart's package manager (e.g., dart pub add schemantic).
  • [REMOTE_CODE_EXECUTION]: In references/genkit_mcp.md, the documentation demonstrates running MCP servers using npx, which involves downloading and executing packages from the NPM registry (e.g., @modelcontextprotocol/server-filesystem).
  • [COMMAND_EXECUTION]: The skill utilizes shell commands for development workflows, including genkit start to wrap Dart execution and dart analyze for code verification.
  • [INDIRECT_PROMPT_INJECTION]: The SkillsPlugin documented in references/genkit_middleware.md allows the agent to read SKILL.md files from local directories and inject their content into the system prompt. This creates a surface for indirect prompt injection if the source directories contain untrusted or externally influenced files.
  • [DATA_EXPOSURE]: The FilesystemPlugin provides the agent with tools to list, read, and write files within a root directory. While intended for workspace management, this represents a significant data access surface that relies on proper root directory scoping for safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 03:05 PM
Security Audit — agent-trust-hub — developing-genkit-dart