printing-press-publish

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and interprets open/public GitHub content — e.g., cloning/refreshing the printing-press-library repo and calling gh pr list, gh pr view --comments, and gh api .../pulls/<PR>/comments — and then uses those PRs/comments and repo state to decide collisions, branch/PR actions, and to instruct the agent to resolve Greptile findings, so untrusted third-party content can materially influence the agent's actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill executes remote code at runtime via "go run golang.org/x/vuln/cmd/govulncheck@v1.3.0 ./..." which fetches the module from golang.org and runs it, making it a required external dependency that executes remote code (golang.org/x/vuln/cmd/govulncheck@v1.3.0).