mcp-builder
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
scripts/evaluation.pyandscripts/connections.pymodules facilitate the execution of local shell commands via thestdiotransport. This behavior is user-driven through command-line arguments and is essential for the skill's documented purpose of testing local MCP servers. - [EXTERNAL_DOWNLOADS]: The documentation and guides within the skill reference official SDKs and protocol specifications hosted on the
modelcontextprotocolGitHub organization, which is a well-known and trusted source for this technology. - [PROMPT_INJECTION]: The evaluation harness in
scripts/evaluation.pyreads question data from an XML file and interpolates it into the prompt sent to the LLM. This provides a surface for indirect prompt injection if the input XML content is provided by an untrusted source, as there are no specific sanitization or boundary markers implemented for the question text.
Audit Metadata