git-pre-release-conflict

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • COMMAND_EXECUTION (SAFE): The skill utilizes standard git CLI commands (git merge, git push, git branch -D) for repository management. All potentially destructive actions like deleting branches or force-pushing include explicit instructions to seek user confirmation first.
  • DATA_EXFILTRATION (SAFE): Network operations are limited to standard git fetch and git push targeting the repository's configured 'origin' remote. No sensitive local files are accessed or transmitted.
  • INDIRECT_PROMPT_INJECTION (LOW): The skill possesses an attack surface for indirect prompt injection as it involves the agent reading and interpreting file contents containing git conflict markers.
  • Ingestion points: git status, git diff, and git show (references/conflict_playbook.md).
  • Boundary markers: None explicitly defined to separate file content from instructions, though the SOP focuses on syntax markers (<<<<<<<).
  • Capability inventory: Branch deletion (git branch -D), merging, and pushing to remotes (references/flow.md).
  • Sanitization: No sanitization of file content is performed before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:23 PM