recur-quickstart
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data such as product names, prices, and API keys which are interpolated into shell commands and MCP tool calls. This creates a surface where malicious input could influence agent behavior.
- Ingestion points: User prompts for product creation and environment setup instructions in SKILL.md.
- Boundary markers: No explicit delimiters are present to isolate user-provided data from instructional context.
- Capability inventory: The skill can execute shell scripts (scripts/check-env.sh), install Node.js packages, and call MCP tools for product and key management.
- Sanitization: No evidence of input validation or sanitization before processing user parameters.
- [COMMAND_EXECUTION]: The skill includes a utility script, scripts/check-env.sh, which inspects the local environment. It reads environment variables and scans .env configuration files to verify the integration setup.
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install official vendor packages (recur-tw, @recur-tw/cli) from npm and connect to a vendor-managed MCP server (https://mcp.recur.tw/). these represent standard vendor resources for the platform.
Audit Metadata