Add Salient to Gemini¶
Connect Salient's MCP server to Google Gemini CLI or Gemini Enterprise.
Gemini CLI (stdio)¶
Gemini CLI supports stdio MCP servers natively. Add Salient to your settings file:
Edit ~/.gemini/settings.json:
Edit %USERPROFILE%\.gemini\settings.json:
{
"mcpServers": {
"salient": {
"command": "python3",
"args": ["/path/to/salient/tools/mcp-ttx/server.py"],
"env": {
"SALIENT_API_URL": "http://localhost:8000",
"SALIENT_TOKEN": "your-jwt-token"
}
}
}
}
Restart Gemini CLI after saving. Verify with:
No token needed for local dev
When running Salient locally with make dev, leave SALIENT_TOKEN empty.
Gemini Enterprise (Google Cloud)¶
For Gemini in Google Cloud (Vertex AI), you need to register a custom MCP connector:
- Open Google Cloud Console > Vertex AI > Gemini
- Navigate to Extensions or Custom Connectors
- Create a new MCP connector with your remote Salient MCP endpoint URL
- Set authentication to Bearer token with your
SALIENT_TOKEN
Remote server required
Gemini Enterprise requires an HTTP MCP endpoint. See the Remote MCP Server guide.
Consumer Gemini Web App¶
The consumer Gemini web app (gemini.google.com) does not currently support custom MCP servers. Use Gemini CLI for local setups or Gemini Enterprise for cloud deployments.
What Works¶
Once connected, Gemini can access all 40+ Salient tools:
- Browse and run tabletop exercises
- Query the digital twin
- Generate posture assessments and playbooks
- Sync identity connectors (Okta, Entra ID)