Skip to content
Apps / Connectors · MCP-native · Plus / Team / Enterprise
PortEden

Secure Your Data With ChatGPT

PortEden makes ChatGPT enterprise-ready in 30 seconds. Add PortEden as a Connector and ChatGPT picks up PII redaction, per-action RBAC, and a SIEM-exportable audit log — so the same ChatGPT your team already uses can finally work on real client data.

See pricing

ChatGPT Plus / Team / Enterprise · 30-second OAuth · Free PortEden tier

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 ChatGPT

  • ChatGPT Connectors expose internal tools to the model. PortEden's Connectors apply redaction and access policy at the request boundary, so the data surfaced to ChatGPT is governed by your rules rather than by the user's underlying OAuth scope.
  • OpenAI's no-training commitment applies to Team and Enterprise; consumer Plus accounts are not covered by default. PortEden's redaction provides defense-in-depth on covered tiers and brings equivalent identifier protection to Plus, regardless of subscription level.
  • Custom GPTs package workflows for reuse. PortEden enforces the firewall on every invocation and records each call under the calling user's token — the same governance model whether the GPT is private, shared with a team, or distributed across the organization.
  • ChatGPT Agents chain tool calls across extended sessions. PortEden produces an end-to-end audit trail covering every tool call in the chain (request, decision, redacted response), so agentic deployments remain reviewable, exportable, and aligned with your retention policy.
OpenAI Privacy Policy
Capability Matrix

ChatGPT With PortEden: What You Get

Real-time PII redaction in tool responses
50+ identifier types stripped before OpenAI receives the response
ChatGPTNo
+ PortEdenYes
Per-tool, per-action RBAC
ChatGPTNo
+ PortEdenYes
Per-contact / per-domain firewall rules
ChatGPTNo
+ PortEdenYes
SIEM-exportable audit log
Enterprise audit logs cover admin events; PortEden adds per-tool-call detail
ChatGPTPartial
+ PortEdenYes
Native MCP via Apps (formerly Connectors)
OpenAI renamed Connectors to Apps in Dec 2025; full MCP support shipped 2025
ChatGPTYes
+ PortEdenYes
Data not used for training
OpenAI guarantee on Team / Enterprise only; Plus is opt-out
ChatGPTPartial
+ PortEdenYes
OAuth token auto-refresh
ChatGPT handles refresh; PortEden inherits
ChatGPTYes
+ PortEdenYes
Token-context reduction (smaller, faster prompts)
~80% reduction via context hygiene
ChatGPTNo
+ PortEdenYes
IdP identity sync (Okta / Entra / Google Workspace)
Enterprise SSO covers the user; PortEden ties firewall rules to IdP groups
ChatGPTPartial
+ PortEdenYes
Confirm-before-write on destructive actions
Send / delete / forward require explicit confirmation
ChatGPTNo
+ PortEdenYes
One-click revocation across surfaces
ChatGPTNo
+ PortEdenYes
MCP / Agentic Coverage

Every ChatGPT Connector, One Firewall

Connection: Native MCP via ChatGPT Apps (formerly Connectors) — 30-second OAuth per endpoint

ChatGPT (Plus / Team / Enterprise)

Settings → Apps & Connectors → Add → paste a PortEden MCP URL. ChatGPT handles OAuth and auto-refresh; PortEden enforces redaction and RBAC on every tool call. (Apps is OpenAI's Dec 2025 rename of Connectors — same MCP under the hood.)

Setup guide

Custom GPTs

Configure a custom GPT to use the PortEden Connector. The GPT inherits your firewall — even if you share it with a team, every call is audited under the calling user's token.

Setup guide

ChatGPT Agents (deep research / browse / agentic)

When ChatGPT chains tool calls in a long agentic loop, every step routes through PortEden. Per-token quotas keep loops bounded; the audit log captures the full chain.

Setup guide
Tooling Notes
  • Each PortEden Connector is added separately in ChatGPT — pick the ones you need (email, calendar, drive, docs, sheets, tasks).
  • ChatGPT auto-refreshes tokens in the background. Users don't see expired-auth errors mid-conversation.
  • Function-calling and tool-use both pass through PortEden's redaction layer — same firewall for ad-hoc chats and agentic workflows.
  • Per-token monthly request quota; exceed it and ChatGPT surfaces a clean "quota exceeded" message.
  • Revoking the token at my.porteden.com immediately stops ChatGPT mid-flight; the next call gets a 401 and ChatGPT prompts for re-auth.

Connect ChatGPT in Under 5 Minutes

1

Sign in to PortEden and connect a data source

OAuth into Gmail, Outlook, Calendar, Drive, or Tasks at my.porteden.com. Pick the access profile that matches your use case.

2

Add the PortEden Connector in ChatGPT

Settings → Connectors → Add connector → paste https://mcp.porteden.com/email (or any other endpoint). Sign in to PortEden via the popup.

3

Use ChatGPT exactly as you do today

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

Read the full ChatGPT guide →
ChatGPT + PortEden

Five-Minute Setup. Free While You Test.

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

Read the ChatGPT guide

Frequently Asked Questions

Do I need ChatGPT Plus, Team, or Enterprise to use the PortEden Connector?
Custom Connectors require a paid ChatGPT plan — Plus, Team, or Enterprise. Free ChatGPT cannot add MCP Connectors. PortEden's free tier is independent of OpenAI's tier — solo licensed practitioners can use the firewall on the free tier with paid ChatGPT.
Does OpenAI get a copy of my data when I use the PortEden Connector?
OpenAI sees the redacted output PortEden returns to the tool call — never the raw email, calendar, or drive payload. Identifiers, free-text PII, and any field your access rules block are replaced with placeholders before the response leaves PortEden's perimeter.
What about training? Will my prompts train the model?
OpenAI's data-not-used-for-training guarantee covers ChatGPT Team and Enterprise tiers. On Plus, you have to opt out per chat. The data ChatGPT pulls through PortEden Connectors — emails, documents, calendar events — is redacted at the firewall before it reaches the model, so even if OpenAI logs the conversation, the protected fields aren't in the tool responses that fed the answer.
Does each ChatGPT Connector require a separate OAuth?
Yes — each PortEden endpoint (email, calendar, drive, docs, sheets, tasks) is a separate Connector with its own OAuth flow. The flow is 30 seconds each. You don't have to add all six; pick the ones the user actually needs.
What happens when ChatGPT tries something the firewall blocks?
The Connector returns a structured permission-denied response. ChatGPT sees the rejection and tells the user ("I tried to read that contact's emails but your firewall blocked it"). The audit log captures the tool call, the arguments, and the rule that fired — no silent failures.
Can a Custom GPT I publish use the PortEden Connector for its users?
No — Connectors are user-scoped. When User A uses your Custom GPT, the GPT calls Connectors using User A's tokens. This is by design: the firewall always audits under the actual user, not the GPT builder. If you need a multi-tenant Connector for a published GPT, talk to sales.
How do I revoke ChatGPT's access?
Sign in to my.porteden.com and revoke the token. The change is immediate — ChatGPT's next tool call gets a 401 and prompts the user to re-authorize. No need to remove the Connector from ChatGPT itself.
Will this slow ChatGPT down?
Redaction adds ~150-200 ms per tool call. Because PortEden also reduces token usage by ~80% via context hygiene, end-to-end latency typically drops on data-heavy tasks. ChatGPT answers faster on PortEden's clean payloads than on a raw API response.

Keep Exploring

Get More From ChatGPT 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 →