speckle-impl-versioning

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow explicitly fetches and ingests user-generated model data from the public Speckle service (e.g., operations.receive with ServerTransport referencing app.speckle.systems and UrlHelper.getResourceUrls / SpeckleLoader loading https://app.speckle.systems/projects/… model URLs in SKILL.md and references/examples.md), which the agent must read/interpret to compute diffs, drive rollbacks, and load viewer content—allowing third-party content to materially influence actions.