Skills
Navigate
skills/chrissargent48/antigravity-awesome-skills/agent-manager-skill/Gen Agent Trust Hub

agent-manager-skill

Fail

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct the user to clone a repository from an external GitHub account.
  • Evidence: git clone https://github.com/fractalmind-ai/agent-manager-skill.git in SKILL.md.
  • [REMOTE_CODE_EXECUTION]: The skill directs the user to run Python scripts from the downloaded external repository, leading to the execution of unverified code.
  • Evidence: python3 agent-manager/scripts/main.py doctor and subsequent commands in SKILL.md.
  • [COMMAND_EXECUTION]: The skill utilizes tmux and python3 to manage local system processes and sessions.
  • Evidence: Common commands and Notes section in SKILL.md referencing tmux and python3 usage.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an interface to assign tasks using arbitrary text input, creating a surface for indirect prompt injection.
  • Ingestion points: The assign command in SKILL.md which uses a heredoc to take input.
  • Boundary markers: Uses heredoc EOF markers, but the content is passed to the underlying script without explicit safety instructions.
  • Capability inventory: The skill has the ability to manage processes via tmux and execute local scripts.
  • Sanitization: There is no evidence of sanitization or validation for the input provided to the assign command.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Nov 1, 2026, 11:35 PM