he4rt-twitter-engagement-tracker

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This extension deliberately intercepts Twitter/X GraphQL responses in the user's browser to harvest tweets, user profiles and engager metadata and includes automatic export/webhook code to push that data to external endpoints, constituting clear data-exfiltration and privacy-abuse capability.

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

• Third-party content exposure detected (high risk: 0.85). The skill patches window.fetch in interceptor.js and then calls clonedResponse.json() on runtime GraphQL responses from https://x.com/* / https://twitter.com/*, extracting legacy.full_text and other tweet/user fields (outsider-authored content) into the extension/background LLM context via forwarded JSON payloads.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)