AI integrations · Gemini
Gemini

Run Manifestly from Gemini.

Manifestly's MCP server connects into Gemini Enterprise and Agent Designer as a custom data source. Your team can ask Gemini to start runs, complete steps, reassign work, and even author new SOPs — all from inside the Google Workspace tools they already use.

Endpoint
mcp.manifest.ly
Protocol
MCP · Streamable HTTP
Auth
OAuth 2.1
Setup
Admin · ~ 10 min
Why Gemini + Manifestly

Workflows that live with your data.

If your team runs on Google Workspace, Gemini is already next to your docs, sheets, calendars, and meetings. Plug Manifestly in as a custom MCP data source and Gemini becomes a workflow surface, not just a Q&A box: it can do the work, not just talk about it.

Build it once in Agent Designer. Distribute it to departments. Each user authenticates with their own Manifestly account — permissions, audit, and SSO behavior carry over cleanly.

  • Works with Gemini Enterprise and Agent Designer
  • Native MCP connector — standard Streamable HTTP
  • OAuth 2.1, scoped per user
  • Full read, write, and author surface
  • Every action attributed in the audit log

Have Gemini set itself up. Seriously.

Copy the prompt below, paste it into Gemini, and Gemini will read these pages, figure out your setup, and walk your admin through the right path. Especially handy because Gemini's connector path varies by Workspace edition.

Copy & paste into Gemini
I want to connect Manifestly Checklists to Gemini using the Model Context Protocol so my team can run workflows by chatting with you.

Please:

1. Read https://www.manifest.ly/integrations/mcp-server and https://www.manifest.ly/ai-integrations/gemini so you understand exactly what Manifestly's MCP server does and how to add a custom MCP data source to Gemini.
2. Ask me whether I'm using Gemini Enterprise directly, building an agent in Agent Designer, or wiring this in for a Workspace tenant — the click path differs by surface.
3. Walk me (or my Workspace admin) through configuring Manifestly as a custom MCP data source with endpoint https://mcp.manifest.ly, and explain how to test the OAuth handoff using a Manifestly account.
4. Once I confirm we're connected, suggest three useful first prompts the team can try (e.g. "what's overdue across operations this week?", "start the new client onboarding workflow for Acme", "draft me a SOC2 evidence collection workflow and import it").
5. Remind me that every action you take via Manifestly is attributed to the human user in the audit log, scoped to that user's existing Manifestly permissions.

If your Gemini surface doesn't have outbound web access enabled, follow the manual setup steps below instead.

What you can ask

Three patterns that turn Gemini into a workflow surface.

Gemini's strength is sitting next to the rest of Workspace — meetings, docs, calendars. These three patterns are where most teams get the most value.

01 / Pre-standup brief

Ask before your morning meeting.

Open Gemini next to your calendar. Ask what's overdue and who's stuck before you walk into standup. Pulls live data, scoped to your permissions.

For: managers · ops leads · team leads
Gemini · Manifestly connected
You
Standup in 10 min. What's overdue on the ops team this week?
Gemini
Pulling overdue assignments for the ops team.
→ list_my_assignments(team=ops, status=overdue)
5 overdue assignments across 4 runs:
• Sam — Q2 SOC2 evidence · "Collect access reviews" · 5d late
• Sam — Q2 SOC2 evidence · "Sign vendor questionnaire" · 3d late
• Priya — Acme onboarding · "Provision SSO" · 2d late
• Jordan — Acme onboarding · "Send welcome packet" · 1d late
• Priya — Northwind kickoff · "Schedule kickoff call" · 1d late
You
Reassign Sam's two to me and add a comment that I'm covering this week.
Gemini
Done. Both steps reassigned to you with the explanatory comment.
02 / In the flow

Run the workflow without leaving Workspace.

Start runs, complete steps, set due dates, capture data fields, leave comments — right from Gemini, while you're already on the doc or thread that prompted the action.

For: anyone executing real work
Gemini · Manifestly connected
You
Start the client onboarding workflow for Northwind, mark step 1 complete, and assign the contract review to Priya.
Gemini
Starting the run, completing step 1, and assigning the contract review.
→ start_run_by_workflow_name("Client onboarding")
→ complete_run_step(run=14228, step=1)
→ assign_run_step_by_name(run=14228, step="Contract review", user="Priya")
✓ Run #14228 — "Northwind onboarding" started
✓ Step 1 "Create CRM record" — completed
✓ Step 4 "Contract review" — assigned to Priya Mehta, due Fri
You
Add a comment on step 7: "Northwind needs SOC2 type 2 — see RFP, page 14."
Gemini
Comment added on step 7. Anything else?
03 / The wow

Build agents that run workflows.

In Agent Designer, build a department-specific agent that talks to Manifestly directly — e.g. an "onboarding concierge" that starts the right run, assigns the right people, and pings the right Slack channel automatically when a new client closes.

For: ops architects building durable internal agents
Agent Designer · Manifestly + Slack connected
You
Build an onboarding-concierge agent: when I say a client closed, start onboarding for them, assign the AE as run owner, and post a kickoff message in #cs.
Gemini
Drafting the agent. It will use Manifestly to start runs and Slack to post the kickoff message.
→ get_workflow_schema()
→ list_workflows(name~="onboarding")
→ register_agent("Onboarding concierge")
✓ Agent published: "Onboarding concierge"
   Trigger: chat command "client closed: <name>"
   Actions: start "Client onboarding" run · assign AE · post #cs message
   Permissions: scoped to caller's Manifestly account
You
Client closed: Northwind. AE is Priya.
Manual setup

Two ways in. Both admin-driven.

Gemini Enterprise wires custom MCP data sources at the Workspace level. Agent Designer lets builders register them per-agent. Both paths use the same endpoint.

Gemini Enterprise

Workspace-wide connector

  1. A Workspace admin opens the Gemini admin console.
  2. Under Data sources → Add custom MCP source, name it Manifestly.
  3. Paste the endpoint URL and configure the OAuth callback per Google's instructions.
  4. Roll out to org units or specific groups. Each user authenticates with their Manifestly account on first use.
https://mcp.manifest.ly
Agent Designer

Per-agent data source

  1. In Agent Designer, open the agent's Data sources panel.
  2. Add a custom MCP source using the endpoint URL.
  3. Test the OAuth handoff with a builder's Manifestly account.
  4. Publish the agent to the intended audience. Each end-user authenticates separately.
https://mcp.manifest.ly

Need IAM details? The MCP server reference documents the OAuth flow, the available tools, and how permissions map from Manifestly to the connected agent.

Built right

Gemini can only do what the user can do.

Enterprise AI integrations have to be auditable. We built this to enterprise standards from day one.

— Auth
OAuth 2.1, no API keys to share.

Users sign in with their Manifestly account on first use. Google never sees a long-lived secret, and admins never paste one.

— Permissions
Scoped per user.

Even when an agent is shared org-wide, each user only sees what their Manifestly account permits. Roles, departments, and assignments carry over.

— Audit trail
Every action attributed.

When Gemini completes a step, the audit log records the human user who asked — not a generic service account.

— Admin control
Revoke at the source.

Admins can revoke the connector tenant-wide from the Workspace admin console. Users can revoke individually from their connected apps page.

Related

Not using Gemini? We have you covered.

Bring Manifestly to Gemini. Start free, or talk to us about your AI strategy.

With Manifestly, your team will Never Miss a Thing.

Dashboard