init-apic-ai-assets

SUSPICIOUS: the stated purpose is coherent, but the reviewed fragment is only a dispatcher and defers real behavior to missing files while relying on external CLI-based skill installation and likely transitive skill installs. Risk is driven by incomplete auditability and supply-chain trust expansion, not confirmed malicious behavior.

The fragment is an installation/configuration guide for API Connect MCP servers. It does not contain clear malware, credential harvesting, or exfiltration logic, but it introduces supply-chain risk by executing unpinned remote content with npx and installing a globally unpinned latest CLI. Persisting client secrets in .bob/mcp.json and permitting TLS validation to be disabled are significant security concerns. Review package provenance, pin versions and hashes, protect the configuration file, and avoid NODE_TLS_REJECT_UNAUTHORIZED=0 unless strictly required.

This is installation and configuration documentation, not apparent malware code. The principal risks are execution of unpinned remote npm/GitHub content, execution of a locally downloaded package without stated checksum or signature verification, plaintext credential storage in .bob/mcp.json, and the option to disable TLS certificate validation. No direct exfiltration, backdoor, obfuscation, or destructive behavior is evident in this fragment.