resolving-merge-conflicts

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The instructions follow industry-standard practices for version control conflict resolution and do not contain any malicious patterns, obfuscation, or exfiltration attempts.
  • [COMMAND_EXECUTION]: The skill instructs the agent to discover and run the project's own automated checks (tests, typecheck, format). This is a standard development workflow and does not involve executing untrusted remote scripts.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content during the resolution process.
  • Ingestion points: Conflicting file content, git commit history, pull request descriptions, and issue tickets (SKILL.md steps 1 and 2).
  • Boundary markers: None explicitly provided; the agent is expected to differentiate between code and conflict markers naturally.
  • Capability inventory: Execution of project-specific shell commands for testing and git operations (SKILL.md steps 4 and 5).
  • Sanitization: None; the skill relies on the agent's reasoning and the project's automated test suite to ensure correctness.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 06:42 AM
Security Audit — agent-trust-hub — resolving-merge-conflicts