maintainer-ready-contribution

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes standard, read-only Git commands in references/cross-repository-git.md (e.g., git remote -v, git branch --show-current) intended for local repository discovery and state verification.\n- [DATA_EXFILTRATION]: The workflow in SKILL.md and references/readiness-checklist.md explicitly instructs users to remove secrets, personal data, and credentials from drafts. It also mandates human approval before any external interaction or submission.\n- [PROMPT_INJECTION]: Instructions in SKILL.md and references/policy-discovery.md direct the agent to respect and prioritize the receiving project's AI and safety policies, providing a framework for responsible automated assistance.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 08:49 PM
Security Audit — agent-trust-hub — maintainer-ready-contribution