agent-multi-repo-swarm

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). This GitHub skill page repeatedly instructs users to run npx ruv-swarm commands (which download and execute npm packages) from a third‑party repository/account that is not a well‑known vendor, creating a real risk that arbitrary/malicious code could be fetched and executed.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow includes GitHub-based operations (e.g., repo discovery/analyze and PR management via the mcp__claude-flow__github_repo_analyze / mcp__claude-flow__github_pr_manage tools) that would read outsider-authored free text from GitHub issues/PRs/comments as part of those actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes runtime commands that fetch and execute remote code (notably "npx ruv-swarm" and cloning remote repositories via "gh repo clone org/$repo"), so external packages/repos fetched at runtime (e.g., npx ruv-swarm and gh repo clone) can execute remote code and thus present a risk.