Warn by default; block when configured. Claude Code configuration is changed without a separate user request, allowing Claude Code to load the newly registered server.
AI security review
scanned · by lpm-firewall-ai
The package's automatic install hook writes an MCP server registration into Claude Code's user configuration. This is an unauthorized install-time change to an AI-agent control surface, so the version is malicious under the review rules.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review
Trigger
npm install of the package.
Impact
Claude Code configuration is changed without a separate user request, allowing Claude Code to load the newly registered server.
Mechanism
The postinstall hook runs the CLI's MCP registration command, which writes a server entry to ~/.claude.json that launches `dr mcp serve dr-zendesk-kb`.
Policy narrative
During npm installation, the postinstall hook runs `dr mcp register dr-zendesk-kb`. The CLI writes that server into `~/.claude.json` with a command that launches `dr mcp serve dr-zendesk-kb`, changing Claude Code's MCP configuration without a separate user request.
AI identified this as a dangerous AI-agent capability at 99.0% confidence with low false-positive risk.
Evidence for policy risk
The package runs scripts/postinstall.mjs automatically after npm install.
The postinstall hook runs `dr mcp register dr-zendesk-kb` without a separate user command.
MCP registration writes a server entry into `~/.claude.json`.
The registered entry launches `dr mcp serve <name>`, enabling Claude Code to load the server.
Evidence against
The flagged credential and network code did not establish an unrelated credential destination; the install-time MCP configuration write independently determines the verdict.
Avoid installing @datarails-ai/dr-cli@1.6.187. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, check your AI agents' configuration and instruction files (the files this report lists) for changes you did not make, and remove them. This finding does not by itself establish malware intent.