cloning-project

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The skill utilizes common Linux shell commands (ls, cp, mkdir, zip, cat) to manage files. These operations are restricted to the local filesystem (/home/claude/, /mnt/user-data/uploads/, and /mnt/user-data/outputs/) and are necessary for the skill's stated goal of exporting project data.
  • [DATA_EXPOSURE] (SAFE): While the skill accesses sensitive project instructions and user-uploaded files, it does so to provide them back to the user in the 'outputs' directory. There is no evidence of data being sent to external or untrusted destinations.
  • [INDIRECT_PROMPT_INJECTION] (LOW):
  • Ingestion points: The skill extracts content from the active context window (project instructions) and reads files from /mnt/user-data/uploads/.
  • Boundary markers: The skill uses markdown and XML markers to identify content for extraction.
  • Capability inventory: The skill has the ability to execute bash commands and write files to the output directory.
  • Sanitization: The use of a quoted heredoc (<< 'INSTRUCTIONS') in Step 2 is a good practice as it prevents the shell from performing variable expansion or command substitution on the extracted content, mitigating basic injection risks during the file-writing process.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:28 PM