draw-io
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and modify
.drawioXML files, creating a surface for potential instruction injection. \n - Ingestion points: The agent reads and processes
.drawioXML files as specified in the SKILL.md and README.md instructions.\n - Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore or isolate embedded instructions within the XML content.\n
- Capability inventory: The skill allows shell command execution (
drawioCLI,pre-commit,mise), Python script execution, and repository modifications (git addin conversion scripts).\n - Sanitization: No validation or sanitization is performed on the XML content before the agent processes the element values.\n- [COMMAND_EXECUTION]: The skill facilitates the execution of external CLI tools including
drawio,pre-commit, andmise, as well as custom shell and Python scripts for icon searching and PNG conversion. This grants the agent broad shell access to interact with the local filesystem and Git repository as part of its normal operation.
Audit Metadata