setup-git-lite

Warn

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [PERSISTENCE_MECHANISMS]: The skill appends environment variable exports to the user shell profile (e.g., .bashrc, .zshrc) to maintain configuration across sessions.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads and modifies global configuration files located in ~/.claude/settings.json, which contains sensitive agent configuration.
  • [COMMAND_EXECUTION]: The skill executes a local Node.js script (setup-git-lite.js) by interpolating subcommands parsed from user arguments directly into a shell command string.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from $ARGUMENTS to determine which subcommand to execute, creating a potential command injection surface.
  • Ingestion points: User input passed via the $ARGUMENTS variable.
  • Boundary markers: No explicit delimiters or ignore instructions are provided for the argument interpolation.
  • Capability inventory: The skill can execute shell commands via node and write to sensitive filesystem locations.
  • Sanitization: No explicit validation or escaping of the subcommand argument is defined in the instructions before it is passed to the shell.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 21, 2026, 02:00 AM
Security Audit — agent-trust-hub — setup-git-lite