linkedin-dm-signal-classifier

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill specification intentionally extracts and forwards users' private LinkedIn DMs (verbatim) to external systems (Slack channels, CRM, enrichment tools) and therefore embodies deliberate data-exfiltration behavior that can leak sensitive information if misconfigured or sent to untrusted endpoints.

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

• Third-party content exposure detected (medium risk: 0.30). Source: LinkedIn DMs of the trigger owner are fetched and then the workflow reads the most recent inbound message text (including verbatim quotation) before posting to Slack, so outsider-authored DM content is ingested at runtime via Steps 2–3 and embedded in Step 6.