ai-automation-workflows
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Example scripts, specifically the data processing template, utilize shell expansion $(cat $file) within command arguments. This pattern is vulnerable to command injection if the processed files contain shell-special characters or malicious payloads designed to escape the string context and execute unintended commands on the host.- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for processing external data that lack typical security guardrails.
- Ingestion points: Data is read from local text files in the data_processing.sh example.
- Boundary markers: Absent; file content is directly concatenated into the prompt string for the AI model without delimiters or instructions to ignore embedded commands.
- Capability inventory: The scripts possess capabilities to run CLI tools (belt), perform network requests (curl), and write to the file system.
- Sanitization: No sanitization or input validation is performed on the data before it is interpolated into the model prompt.
Audit Metadata