resolving-merge-conflicts

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to identify and execute local automated checks, including type checkers, tests, and formatting tools found within the project repository.
  • [PROMPT_INJECTION]: The agent is directed to ingest data from external sources such as commit messages, pull request descriptions, and issue trackers to understand code changes. This creates a surface for indirect prompt injection where untrusted text could influence agent actions.
  • Ingestion points: Git commit history, pull request metadata, and issue tracker content (SKILL.md).
  • Boundary markers: The instructions lack explicit boundary markers or warnings to disregard instructions found within the processed metadata.
  • Capability inventory: The skill allows the agent to modify the codebase through conflict resolution, stage files, commit changes, and execute local scripts (automated checks).
  • Sanitization: There is no mention of sanitization or filtering of the content retrieved from external repository sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 08:00 AM
Security Audit — agent-trust-hub — resolving-merge-conflicts