write-like-meng-on-x

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow refreshes and expands X-authored posts via the in-app browser (https://x.com/MengTo and https://x.com/search?q=from%3AMengTo&src=typed_query&f=live), then ingests the captured tweet text into references/tweet-corpus.jsonl and later uses that corpus as voice evidence for drafting.