continuous-learning-v2

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

• Third-party content exposure detected (high risk: 0.80). The skill's CLI explicitly fetches and parses arbitrary HTTP(S) URLs in scripts/instinct-cli.py (cmd_import), importing untrusted web-hosted instinct files which are then written into the agent's instincts (and can change future tool behavior and promotions), enabling indirect prompt-injection from third-party content.