parallel-dev-worktrees

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

• Third-party content exposure detected (high risk: 0.80). The skill requires and instructs the agent to consult and/or install the external Portless skill (see SKILL.md and README: "Portless is a required companion" and the explicit npx/GitHub URL and "defer to the official Portless skill: https://github.com/vercel-labs/portless/blob/main/skills/portless/SKILL.md"), which clearly directs the agent to fetch and interpret public GitHub-hosted content that could alter command/installation behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The SKILL explicitly requires installing and invoking the Portless skill at runtime (e.g., via "npx skills add https://github.com/vercel-labs/portless --skill portless" and references https://github.com/vercel-labs/portless/blob/main/skills/portless/SKILL.md), which would fetch and execute remote code as a required dependency for the workflow.