Skip to content
Vertex AI · Workspace · MCP · Function-calling
Google GeminiPortEden

Secure Your Data With Gemini

PortEden makes Gemini ready for the work your team actually needs done. Native MCP servers, Vertex AI agents, function-calling, and the Workspace side panel all get an audited, redacted data layer — so Gemini answers questions you couldn't safely ask it before.

See pricing

Vertex AI · MCP · Workspace add-ons · 5-minute setup

Works With What You Already Use
Gmail
Gmail
Outlook
Outlook
Google Calendar
Google Calendar
Google Drive
Google Drive
Google Docs
Google Docs
OneDrive
OneDrive
SharePoint
SharePoint
Teams
Teams
Slack
Slack
Notion
Notion
Asana
Asana
Monday
Monday
Linear
Linear
Jira
Jira
Confluence
Confluence
Entra ID
Entra ID
Better Together

What PortEden Adds to Gemini

  • Google announced native MCP support for Gemini in 2026, with managed remote MCP servers on Vertex AI and a unified MCP layer across Google services. PortEden registers as a first-class MCP server, so the same firewall that protects Claude and ChatGPT now protects Gemini agents end-to-end.
  • Vertex AI gives you direct control over the inference path. PortEden enforces redaction and policy on every tool call your agents make, so the data Vertex receives via your tools — across MCP, function-calling, Vertex AI agents, and the Workspace side panel — is governed by a single consistent layer.
  • Workspace Gemini grounds inside Google's tenancy by default. PortEden adds firewall-side rules at the folder, contact, and label level, so Gemini grounds only on the resources your policies have explicitly approved.
  • Vertex AI Agent Builder and the Gemini API both consume the PortEden surface — MCP servers, function tools, or the REST API. Every agent runs under per-token RBAC with redacted inputs and an exportable audit trail across every action it takes.
Capability Matrix

Gemini With PortEden: What You Get

Real-time PII redaction in tool responses
50+ identifier types stripped before data reaches the model
GeminiNo
+ PortEdenYes
Per-tool, per-action RBAC
GeminiNo
+ PortEdenYes
Per-contact / per-domain firewall rules
GeminiNo
+ PortEdenYes
SIEM-exportable audit log
Google's audit is split across surfaces
GeminiPartial
+ PortEdenYes
Native MCP support
Google launched official MCP support for Gemini / Vertex AI in 2026
GeminiYes
+ PortEdenYes
Function-calling tool integration
PortEden CLI as a function tool
GeminiYes
+ PortEdenYes
Vertex AI agent integration
Bearer token against PortEden REST API
GeminiYes
+ PortEdenYes
Workspace side-panel data control
Apps Script bridge to PortEden
GeminiPartial
+ PortEdenYes
Token-context reduction (smaller, faster prompts)
~80% reduction via context hygiene
GeminiNo
+ PortEdenYes
IdP identity sync (Okta / Entra / Google Workspace)
Permissions revoke with the rest of the user's accounts
GeminiNo
+ PortEdenYes
Confirm-before-write on destructive actions
Send / delete / forward require explicit confirmation
GeminiNo
+ PortEdenYes
One-click revocation across surfaces
GeminiNo
+ PortEdenYes
Tool-Use Coverage

Every Gemini Surface, One Firewall

Connection: Native MCP + Function-calling + CLI + Apps Script

Gemini / Vertex AI MCP servers

Google's 2026 MCP rollout means Gemini and Vertex AI consume MCP servers directly. Point Gemini at the PortEden MCP server URL — redaction, RBAC, and audit apply to every tool call, no code changes.

Setup guide

Gemini API / Google AI Studio (function-calling)

Declare PortEden CLI commands as function tools. Gemini calls them; your code runs the porteden binary; redacted JSON flows back. Same model whether you use Gemini 3 Pro or Gemini Flash.

Setup guide

Vertex AI Agents

Point a Vertex agent's HTTP tools at api.porteden.com with a Bearer token. Every tool call inherits the token's RBAC, redaction, and audit — no per-agent code changes.

Setup guide

Workspace Gemini side panel (Apps Script)

Bind an Apps Script project to a Workspace add-on. UrlFetchApp.fetch into PortEden returns redacted Workspace data — Gemini grounds on the firewall, not raw Drive.

Setup guide
Tooling Notes
  • Google's 2026 MCP launch makes Gemini a first-class MCP client — PortEden registers as a server and Gemini consumes it the same way Claude or ChatGPT do.
  • Function-calling remains fully supported — PortEden CLI commands map 1:1 to Gemini function declarations for teams not yet on MCP.
  • Vertex AI Agent Builder accepts HTTP tool endpoints; pointing at api.porteden.com gives you a fully audited, per-token-permissioned agent.
  • Workspace Gemini side-panel grounding can be redirected through Apps Script + PortEden — useful for compliance tenants where the default grounding behavior is too broad.
  • Per-token monthly quotas, contact-level access rules, and field-level redaction all apply uniformly across Gemini surfaces.

Connect Gemini in 5 Minutes

1

Generate a PortEden API key

At my.porteden.com, create a token scoped to the data Gemini needs and the actions you want it to take.

2

Wire the token into Gemini

For Gemini API: declare PortEden CLI commands as function tools. For Vertex: paste the token into your agent. For Workspace: store in Apps Script Properties.

3

Use Gemini on real data

Ask Gemini about anything in your stack. Every tool call routes through PortEden — redaction, RBAC, and audit happen automatically.

Gemini + PortEden

Five-Minute Setup. Free While You Test.

Connect a data source, plug Gemini into PortEden, and put Gemini to work on the data your team actually needs to handle.

Frequently Asked Questions

Does Gemini actually support MCP, or is this a custom integration?
MCP is native. Google announced official MCP support for Gemini and Vertex AI in 2026, with managed remote MCP servers and a unified MCP layer across Google services. Gemini consumes MCP servers the same way Claude and ChatGPT do — PortEden registers as one of those servers. Function-calling is still fully supported if your stack hasn't migrated yet; PortEden CLI commands map 1:1 to Gemini function declarations.
Will Gemini for Workspace work with PortEden on the side panel?
Yes — via the Apps Script bridge. Bind an Apps Script project to a Workspace add-on, expose functions that call PortEden's REST API, and the Gemini side panel can invoke them. The user sees Gemini grounding on Workspace data, but the data has already been redacted by PortEden.
Does this work with Vertex AI Agents and Agent Builder?
Yes. Point Vertex AI Agent tools at api.porteden.com with a Bearer token. Every tool the agent calls passes through PortEden's redaction and RBAC layer. The same token's audit log captures the agent's full reasoning chain.
Does Google get a copy of my Drive data through Gemini?
Workspace Gemini stays inside Google's tenancy by default. The risk is in what tool calls return to the model — share-linked files, sensitive fields in Docs, identifying detail in Drive metadata. PortEden's job is to redact those fields in the tool response before they reach the model, regardless of which Google surface is calling.
Does PortEden support gemini-flash and gemini-2.0-pro both?
Yes — model choice is on the Gemini side. PortEden returns the same redacted JSON regardless of which Gemini variant calls it. Use Flash for cheap high-volume tool calls and 2.0 Pro when you need the bigger context window.
What happens when Gemini's tool call gets blocked by an access rule?
The CLI or REST endpoint returns a structured permission-denied response. Gemini sees the rejection and surfaces it in the answer ("I tried to read that file but your firewall blocked it"). The audit log captures the tool call, the arguments, and the rule that fired.
Is there a free tier?
Yes — free for solo licensed practitioners and small developer use. Higher monthly tool-call quotas, SSO, and SIEM export are on paid plans. See pricing for details.

Get More From Gemini With PortEden

Five-minute setup. Free tier for solo licensed practitioners. Same AI you already use — now ready for the work your team actually needs to do.

Talk to sales

Rolling out to a whole team? Talk to sales →