vectorforce@0.3.8
Portable engineering context platform - Git for AI context
AI security review
scanned · by lpm-firewall-aiThe OpenAI provider accepts ANTHROPIC_API_KEY as a fallback credential, and the client sends the selected key to the OpenAI-configured endpoint. This is cross-service credential disclosure when the user invokes LLM reasoning without specifically authorizing use of the Anthropic key there.