land-and-deploy

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly reads and interprets user-generated, third-party content (GitHub PR bodies, checks and runs via gh commands such as gh pr view --json body and gh pr checks, and external site content via curl / $B goto during canary checks) and uses that content to make merge/deploy decisions in the workflow (e.g., Step 3.5 PR body accuracy and Step 7 canary verification), so it clearly ingests untrusted external content that can influence tool actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's SETUP step will download and execute the Bun installer script at https://bun.sh/install (curl -fsSL "..."/bash) to build the browse tool at runtime, which fetches and runs remote code required for the skill's browsing/deploy verification.