branche

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands git checkout, git pull, and git branch to perform repository operations. These commands are executed locally but rely on variables that can be influenced by user input.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (specifically command injection) because it incorporates untrusted data into a shell command execution environment.\n
  • Ingestion points: The $ARGUMENTS variable in SKILL.md captures user input intended for the branch name and passes it to the underlying system shell.\n
  • Boundary markers: There are no boundary markers, delimiters, or instructional warnings to prevent the agent from interpreting shell metacharacters within the provided argument.\n
  • Capability inventory: The skill possesses the capability to execute arbitrary commands via the shell whenever it processes the git checkout -b instruction.\n
  • Sanitization: The skill lacks explicit sanitization or validation logic to ensure the $ARGUMENTS variable contains only valid branch name characters, allowing for potential command chaining (e.g., via ; or &&).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 03:05 PM
Security Audit — agent-trust-hub — branche