skills/seifip/kilobat-skills/kilobat/Gen Agent Trust Hub

kilobat

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill initiates the installation of external components from the vendor's infrastructure, including a CLI tool via Homebrew (brew install seifip/kilobat/kilobat) and additional skill modules using npx (npx skills add seifip/kilobat-skills).
  • [COMMAND_EXECUTION]: The skill extensively utilizes a command-line interface (kilobat and its alias kb) to perform network analysis and system management tasks. It includes explicit instructions to perform safety checks and obtain user authorization before executing potentially disruptive or system-altering commands.
  • [CREDENTIALS_UNSAFE]: The tool includes capabilities to retrieve and share saved network credentials (wifi password, wifi share). These operations are explicitly gated behind safety instructions that require a specific user request and a macOS system prompt.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input such as URLs and hostnames through the diagnose_connection tool. This represents an attack surface where malicious data from external sites could attempt to influence agent behavior, though the skill provides structured safety rules to mitigate operational risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 01:23 PM