skills/ortus-boxlang/skills/bx-ftp/Gen Agent Trust Hub

bx-ftp

Warn

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXPOSURE]: The skill is designed to interact with sensitive filesystem locations, specifically referencing SSH private keys (e.g., /home/app/.ssh/id_rsa) to facilitate SFTP authentication.
  • [DATA_EXFILTRATION]: The skill enables the outbound transfer of local files to external remote servers via the putFile action, which could be leveraged to exfiltrate sensitive information.
  • [COMMAND_EXECUTION]: The skill documentation includes shell commands for environment setup, such as install-bx-module bx-ftp and box install bx-ftp, which are standard installation procedures for the BoxLang ecosystem.
  • [INDIRECT_PROMPT_INJECTION]: The skill functions as a data ingestion point by fetching file listings and file contents from external, potentially untrusted FTP/SFTP servers.
  • Ingestion points: Data returned from listDir and getFile actions in SKILL.md are processed by the agent.
  • Boundary markers: There are no defined delimiters or instructions to prevent the agent from interpreting remote file content or metadata as commands.
  • Capability inventory: The skill possesses extensive capabilities including local file writing (getFile), network communication (open, putFile), and directory manipulation (createDir, deleteDir).
  • Sanitization: The skill lacks explicit instructions for sanitizing or validating data retrieved from remote servers before it is used in subsequent agent logic.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 8, 2026, 09:47 PM