git-expert

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a wrapper for executing Git CLI commands on the local system. \n
  • Evidence: The script scripts/main.cjs uses child_process.spawn to run the git binary with arguments passed from the agent. \n
  • Security Control: The implementation uses shell: false and windowsHide: true, which effectively mitigates common shell injection vulnerabilities by preventing the command from being interpreted by a shell environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves reading and interpreting data from Git repositories, which creates a potential surface for indirect prompt injection through malicious repository content. \n
  • Ingestion points: Git command outputs from operations like git status, git diff, and git log as seen in SKILL.md and scripts/main.cjs. \n
  • Boundary markers: The current implementation does not use explicit delimiters or specific instructions to the agent to ignore instructions embedded within the Git output. \n
  • Capability inventory: The skill has the capability to execute any Git command, which includes file system modification and network operations (e.g., git push, git fetch). \n
  • Sanitization: No evidence of output sanitization or filtering was found in the provided scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 12:18 AM