The package has no install hook or established credential-exfiltration behavior in the reviewed paths. However, MCP generation automatically edits Claude Desktop’s broader server configuration, including through a rate-limit fallback, which presents a real agent-capability risk.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review
Trigger
Running the `mcp` or `skill` command, or a Claude task reaching the rate-limit fallback.
Impact
Claude Desktop may launch the generated server in future sessions; this is a broader agent setting change that merits user awareness.
Mechanism
The builder adds an entry under `mcpServers` and writes the Claude Desktop configuration without a separate registration prompt.
Evidence
cli.jssrc/pilot/mcp/mcp-skill-builder.jssrc/pilot/adapters/claude-automation-adapter.js%APPDATA%/Claude/claude_desktop_config.json