awaken

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s required runtime workflow ingests outsider-authored free text directly from the user in “Phase 1: รู้จักกัน — Batch Freetext” (a single message containing freetext that is parsed into oracle_name/human_name/purpose/theme_hint/etc.), making the agent respond to arbitrary user-provided text without pre-selecting a specific item.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill runs /learn against external GitHub repos (https://github.com/Soul-Brews-Studio/opensource-nat-brain-oracle and https://github.com/Soul-Brews-Studio/oracle-v2) during the Full Soul Sync to fetch repository content that the agent uses to discover and generate its philosophy and files, meaning remote content fetched at runtime directly influences the agent's prompts/output.