constructive-boilerplate-authoring

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly loads and uses boilerplate repositories from arbitrary public repos (e.g., the default constructive-io/pgpm-boilerplates and user-specified repos via "pgpm init --repo myorg/my-boilerplates"), ingesting .boilerplate.json and template files from third-party sources which are parsed to drive prompts and scaffolding behavior and thus could enable indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill fetches templates at runtime from an external boilerplate repository (default: constructive-io/pgpm-boilerplates), and those repository files (.boilerplate.json templates) directly define the prompts/questions used by pgpm, so the remote repo controls agent/user prompts.