mcp-server-builder
Warn
Audited by Gen Agent Trust Hub on Apr 5, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to run scripts located in a 'scripts/' directory (e.g., 'python3 scripts/openapi_to_mcp.py'). These commands are defined in the 'Key Workflows' section of SKILL.md.
- [REMOTE_CODE_EXECUTION]: The executable scripts mentioned in the workflows are missing from the provided skill bundle. Executing unverified local code is a security concern as the contents and behavior of these scripts cannot be audited.
- [PROMPT_INJECTION]: The skill processes untrusted OpenAPI specifications, creating a vulnerability to indirect prompt injection. 1. Ingestion points: External files (openapi.json) and stdin. 2. Boundary markers: No explicit delimiters or safety instructions are defined to separate untrusted data from the agent's task context. 3. Capability inventory: The skill has the capability to write files to the system and execute shell scripts. 4. Sanitization: Although the guide mentions snake_case sanitization for names, the core processing logic is unverifiable due to the missing scripts.
- [SAFE]: The skill follows security best practices for credential management by explicitly forbidding hardcoded secrets and recommending the use of environment variables in its generated templates.
Audit Metadata