gitbutler

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): Indirect Prompt Injection surface identified in the Claude Code integration hooks documentation.
  • Ingestion points: The 'but claude stop' hook processes workspace state and context (untrusted repository data) to generate commit messages (references/hooks.md).
  • Boundary markers: The suggested configuration does not include explicit boundary markers or instructions to ignore instructions embedded in the code being committed.
  • Capability inventory: The integration enables automated git commits ('but commit') based on AI-generated content.
  • Sanitization: No sanitization or validation of the ingested context is specified in the configuration guide.
  • SAFE: The CLI commands provided in references/cheatsheet.md and references/tutorial.md are standard, non-malicious usage examples for the GitButler tool.
  • EXTERNAL_DOWNLOADS (SAFE): External URLs in the documentation point to official resources (gitbutler.com, anthropic.com) and do not involve remote code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:50 PM