Integrations

Model providers, your code host, your inbox and calendar, your cloud accounts, a real browser, and any MCP server you add. Each connection is granted per agent and every use is logged.

Any model, per agent.

Connect the providers you already pay for and match each agent to the model that suits its role. If one provider fails, the next on the agent’s failover list is tried.

Anthropic

Claude

Cohere

Command

Google

Gemini

Free access

Groq

Fast open-weight inference

Mistral

Mistral & Codestral

OpenAI

GPT

OpenRouter

Hundreds of models, one key

Free access

Together AI

Open models

xAI

Grok

Pay as you go

Custom

Any OpenAI-compatible endpoint, or models running on your own environment.

GitHub, or version control you host.

Agents clone the repository, work on a shared task branch, open the pull request, read CI and merge, with every commit attributed. You can also run version control on your satellite, and Cequenced Team can set it up for you.

  • One branch per objective; subtasks accumulate into a single pull request.
  • Merging is a granted capability, so only agents you allow can do it.
  • CI status is visible next to every task.
Agent activity feed showing commits, merges and approvals
Commits, merges and approvals on the record

Gmail and Google Calendar

Agents you permit can read and send mail, run outreach sequences and schedule meetings. Each action is logged against the person who authorised it.

  • Revoke access from either side at any time.
  1. 06:12Sales Repsent step 3 to 14 prospects
  2. 06:15Sales Repreply from acme.co, escalated to you
  3. 06:40Assistantbooked Thursday 10:00 with Priya
  4. 07:31Marketerlaunch email drafted, waiting for your edit
Example: mail and calendar in the activity feed

AWS and Google Cloud

Give agents scoped AWS credentials and a Google Cloud service account. The Budget dashboard records your actual AWS and Google Cloud spend and your actual billing from Anthropic, OpenAI, OpenRouter and xAI, and updates the figures as providers finalize them. These are billing figures, not token estimates.

Cequenced Team Usage page showing token consumption by model source, model, role and agent
The Usage pageToken consumption by model source, model, role and agent

The tool catalog

Each tool in the catalog exposes individual capabilities, so you decide what an agent can do: a researcher reads, a coordinator sends, a developer merges. Credentials are never shown to the agent.

Sales & marketing

  • Mailgunsend campaigns, manage lists, read stats
  • Browser Automationresearch any site in a real browser session

Productivity

  • Gmailsearch, read, draft, send, label
  • Google Calendarlist, search, create and update events

Code

  • GitHubcommit, open PRs, read CI, merge
  • Satellite Gitthe same, on your satellite
  • AtlassianJira issues and Confluence pages

Cloud

  • AWSthe full AWS API, read-only by default
  • Google CloudBigQuery, Cloud Storage, Cloud SQL, Compute Engine, Firestore, Pub/Sub, Spanner

Infrastructure

  • PortainerDocker environments, stacks, containers
  • pfSensefirewall rules, interfaces and services
  • Fleet Opsinfrastructure administration in your own environment, with approval before changes

IoT

  • Home Assistantdashboards, devices, areas and automations

Bring your own

  • Any MCP serverconnect your own tools, granted individually
  • Any OpenAI-compatible modelpoint an agent at your own endpoint

Connect your first tool.

Missing a tool you need? Tell us.

Get started free →See pricing