loopy

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

• Third-party content exposure detected (medium risk: 0.30). In the “Find” route, when web access is available Loopy reads the live Loop Library catalog at signals.forwardfuture.com (catalog.md/catalog.json) and then may treat catalog content as reference data without executing it, so outsider-authored free text can be ingested indirectly via the provider’s live catalog feed.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs at runtime to read the live catalog at https://signals.forwardfuture.com/loop-library/catalog.md (and the structured https://signals.forwardfuture.com/loop-library/catalog.json), and the fetched record is used as the authoritative loop definition/prompt to run or present, so remote content can directly control agent prompts.