Support

Agent Integrator is the secure middleware that lets AI agents act on the platforms you already use — Google Calendar, HubSpot, Calendly, GoHighLevel, Zoho, SendGrid, Twilio, Shopify, Stripe, and more. Instead of pasting API keys and refresh tokens, you click "Connect," log in with the provider, and you're done.

You clicked a "Connect" button inside an AI agent platform that bundles Agent Integrator. The provider shows our app name on the consent screen so you can review which permissions you're granting before approving. After approval, you're returned to the platform's dashboard with the integration live.

Only what's needed for the integration. For Google Calendar, that's read-only profile/email plus the ability to read your free-time and create calendar events. For HubSpot, that's read/write on contacts and deals. The exact list shows on the provider's consent screen before you approve. We never request "all-account" or admin scopes when a narrower scope would do.

Access tokens and refresh tokens are encrypted with AES-256 before they touch our database. Plaintext is decrypted only at the moment your AI agent makes a call, and is never written to logs or backups. See the Privacy Policy for full detail.

Open your AI agent platform's integrations page and click Disconnect on the integration you want to remove. We immediately revoke the token at the provider (where revocation is supported) and wipe the encrypted token from our database.

You can also revoke our app from the provider's own security settings — for example:

• Google: myaccount.google.com → Security → Third-party apps with account access
• HubSpot: Account settings → Integrations → Connected apps
• Calendly: Account → Integrations & apps
• Shopify: Apps → Manage private apps / installed apps

Revocation from either side takes effect at the next request.

Tokens normally refresh themselves automatically. If you see a connection error, the most common causes are:

• You revoked the app from the provider's security settings — re-click Connect to re-authorize.
• Your provider account password was changed and they invalidated all sessions — re-click Connect.
• The provider rotated their API or deprecated a scope we depend on — please contact support so we can investigate.

"Reconnect" is always safe — it issues fresh tokens and replaces the old encrypted record.

No. Customer data flowing through connected integrations is used only to fulfill the action your AI agent is making in real time. It is not used to train, fine-tune, or evaluate AI models. The AI platform that bundles Agent Integrator into its product has its own data-handling policy, which governs what your agent does with the data once retrieved.

Probably — we add new providers each release cycle. The criteria are: (1) the provider supports OAuth 2.0 or a clear API-key flow, (2) there's enough customer demand to justify the integration, and (3) the provider's terms allow third-party AI agent use. Email support@agentintegrator.cloud with the provider name and what you'd want the AI agent to do.

For some providers, yes — the platform that bundles Agent Integrator may expose a "manual setup" option that lets you paste your own client_id, client_secret, and refresh token. That path bypasses the gray-label OAuth flow but lets you keep brand control on the consent screen. Ask your platform operator if they expose this option.

Token data is hosted in the region selected by the platform operator that bundles Agent Integrator into their product. The exact region is documented by that platform. If you need details for compliance (SOC 2, GDPR data-residency, HIPAA), reach out to your platform operator first; they can route the question to us if needed.

Account deletion is handled by the AI agent platform that hosts Agent Integrator. When you delete your account there, all associated Agent Integrator tokens are wiped within the standard retention window (encrypted backups overwrite on a 30-day rolling basis). If you'd like confirmation that deletion has occurred, email support@agentintegrator.cloud and we'll verify and respond.

Email support@agentintegrator.cloud with "SECURITY" in the subject line and a clear reproduction. Please don't post details publicly until we've had a chance to investigate. We acknowledge reports within 48 hours and aim to remediate confirmed issues promptly.