Ffind

Warn

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the ffind tool, explicitly requiring sudo for extraction tasks. This represents a privilege escalation risk, as it grants the tool root-level access to the system.
  • [PROMPT_INJECTION]: There is an indirect prompt injection risk. Ingestion points: User-supplied file paths are used directly as arguments for the ffind command in SKILL.md. Boundary markers: The skill does not define delimiters for user input. Capability inventory: The skill performs subprocess execution of a binary. Sanitization: There is no evidence of input validation or escaping for shell metacharacters in the instructions.
  • [EXTERNAL_DOWNLOADS]: The skill depends on external packages (e2fsprogs, f2fs-tools, util-linux) and a binary from the iothackbot project. While these are presented as environment requirements, they are external to the skill's own code and their integrity is assumed.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 26, 2026, 01:00 AM