Trust center

Security, privacy, and human control—without borrowed badges.

This page describes controls visible in the current application source, the data flow they protect, and the limits a company should know before approval. Last reviewed August 1, 2026.

Current controls

Authentication and sessions

Local passwords are salted and hashed with scrypt. Session cookies are HttpOnly, SameSite=Lax, Secure in production, and backed by HMAC-derived server-side session identifiers. Password changes can revoke every active session.

Secrets and customer credentials

Customer-supplied provider keys and connector/webhook secrets use AES-256-GCM envelope encryption with a fresh random IV and authentication tag. Plaintext credentials are not stored by these code paths and only the last four characters are shown.

Tenant, team, and API boundaries

Existing sites migrate into an organization-backed workspace on first use. Owner, admin, sales, analyst, and viewer roles gate site, pipeline, proof, retention, and membership actions; invitations are exact-email and single-use. Public API credentials use least-privilege scopes, paid-entitlement checks, request IDs, rate limits, and idempotency controls.

Signed outbound integrations

Developer webhooks are signed with HMAC-SHA256 over the exact timestamp and body, delivered through a durable retry queue, and blocked from private or internal network destinations to reduce SSRF risk.

Abuse and spend controls

Public routes use capped JSON bodies and per-route rate limits. Agent traffic also supports owner-funded spend ceilings. Obvious bots are excluded from visitor analytics.

Deletion and retention

Owners can export lead data and erase lead PII. Erasure clears identity, message, and follow-up content while preserving non-identifying aggregate facts. An optional deployment retention window can enqueue the same erasure path automatically.

Data flow

  1. 01

    A site owner supplies business content, settings, and optional provider credentials.

  2. 02

    A customer opens a hosted site or embedded widget and sends text, voice, request, booking-request, or handoff data.

  3. 03

    Only the data needed for the configured feature is sent to the selected AI, speech, messaging, email, billing, or infrastructure provider.

  4. 04

    Make Agent Fast stores the site, conversation, request, attribution, optional opportunity, sync, and audit records needed to operate the selected workflow.

  5. 05

    The owner can review activity, export requests, erase PII, or connect signed webhooks to another company system.

Provider categories

Depending on deployment and enabled features, data may be processed by configured hosting/database, AI model, embeddings, speech, messaging, email, billing, domain, and monitoring providers. Supported code paths include OpenAI, Google Gemini, Anthropic, Deepgram, ElevenLabs, Telegram, Meta platforms, Discord, KakaoTalk, Mailjet, Polar, Cloudflare, and PostgreSQL hosting. This is a capability list, not proof that every provider is active in a given deployment. Request the deployment-specific list for procurement.

Procurement honesty

Current limitations

These are not hidden in fine print. They define what a buyer should verify or contract for.

  • No SOC 2, ISO 27001, HIPAA, PCI DSS, or other certification is claimed on this page.
  • Workspace roles and invitations are available. SSO/SAML, SCIM, and enterprise identity lifecycle automation are not yet generally available.
  • Native HubSpot synchronization is planned, not available. CRM-neutral API and signed webhook foundations are available.
  • Database and transport encryption depend on the production hosting and database configuration; request deployment-specific evidence before procurement approval.
  • AI outputs can be wrong. Buyers are told they are interacting with AI, and important commercial, legal, security, and technical details should be confirmed by a person.
  • No incident-response SLA, uptime SLA, DPA, or subprocessor commitment is implied unless it appears in a signed customer agreement.

Need a deployment-specific security review?

Email security@makeagent.fast or include the review in a pilot request.