Anthropic Connector Submit
Security checks across malware telemetry and agentic risk
Overview
This is an instruction-only checklist for submitting an MCP server to Anthropic’s connector directory, with no code execution or hidden behavior found.
Use this as a checklist, but review every external form field before submitting. Share only a dedicated, least-privilege test account with sample data, avoid production secrets, and select privacy, retention, encryption, and GDPR statements only if they are true for your deployed server.
SkillSpector
By NVIDIA
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
VirusTotal
65/65 vendors flagged this skill as clean.