graphify

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The content mandates dispatching external "Agent" subagents with general-purpose (write + bash) privileges to read/process corpus files and to write absolute paths, creating a clear avenue for remote code execution and data exfiltration; it also instructs installing a suspicious/alternate package name (graphifyy) which raises supply-chain concerns.

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text can be ingested into the LLM context via Step 3 Part B semantic extraction: the workflow dispatches an Agent subagent with FILE_LIST pointing to docs/papers/images that may originate from outsider sources (e.g., /graphify <url> / /graphify add <url> fetching webpages/tweets/arXiv/PDFs), and the subagent reads those file contents to produce JSON.