filesystem

Warn

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the Bash tool for potentially destructive operations, such as 'rm -r' for recursive directory deletion and 'chmod' for modifying file permissions. These commands are executed without a sandbox, posing a risk to the host system if the agent is misled.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes file contents using the 'Read' and 'Grep' tools without implementing boundary markers or sanitization logic.
  • Ingestion points: File contents are read into the agent's context through the 'Read' and 'Grep' tools as described in SKILL.md.
  • Boundary markers: No specific delimiters or instructions are provided to help the agent distinguish between its own instructions and potentially malicious commands embedded within the files it processes.
  • Capability inventory: The agent possesses high-impact capabilities including file writing ('Write', 'Edit'), deletion ('rm'), and permission changes ('chmod'), which could be triggered by instructions found in user-controlled files.
  • Sanitization: The skill does not mention any methods for validating, filtering, or escaping file content before it is processed by the language model.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 7, 2026, 08:09 PM