Ir al contenido

Security & governance by plan

Esta página aún no está disponible en tu idioma.

import { Aside } from ‘@astrojs/starlight/components’;

Every ShareOut plan runs on the same security foundation — moderation, encryption, and visibility controls apply to Personal and Teams alike. Teams adds governance on top: internal approval gates, an audit trail, and workspace-scale roles. This page is the matrix; each linked guide has the full detail.

Two independent role systems, at two scopes:

ScopeRolesApplies to
Workspaceowner > admin > memberTeam-wide settings, billing, schedules, automations — see Workspace admin
Artifactowner / editor / viewerA single artifact’s edit and view rights — see Publishing artifacts → Collaborators

External people (clients, partners, suppliers) are a separate member_class — they never appear on your workspace member list, never count toward seats, and only see what’s explicitly shared with them. See External sharing.

ControlFree / Pro (Personal)Teams / Enterprise
AI moderation on public contentYesYes
Credential encryption at restYesYes
Visibility model (private / workspace / public)YesYes
Workspace roles (owner/admin/member)Yes
Publish approvals (require_approval policy)Yes
Audit logYes
Members per workspace3 maxUnlimited
External sharing (clients/partners)Yes

Workspace owners and admins set a public_publish_policy:

PolicyBehaviour
allowDefault. Members publish to open visibility freely.
prohibitMembers cannot go open; artifact stays workspace-visible.
require_approvalHeld at workspace until N nominated approvers approve (1–10).

Approval is pinned to the artifact’s content hash — if the HTML changes after approval, the next publish needs a fresh round. Re-publishing unchanged content does not. Platform AI moderation still runs on top once approved.

Full flow: Publishing artifacts → Workspace publish governance.

An append-only, per-workspace trail, visible to admin/owner:

GET /v1/workspaces/{id}/audit?limit=100

Covers member add/remove, bulk invite, ownership transfer, visibility changes, subdomain enable/disable, membership-policy changes, connection create/delete, and session-policy changes. Retained for one year.

Full reference: Workspace admin → Audit log.

Every artifact that goes to public visibility — on any plan — passes an automated safety review before it serves:

  • An AI content classifier checks for phishing, malware, scams, and illegal content.
  • A URL-reputation check scans outbound links and embedded resources.
OutcomeResult
ApprovedGoes live immediately.
PendingHeld private until human review — the fail-safe: anything the classifier can’t confidently clear, or any check error, lands here rather than going live.
BlockedNot served.

This is a platform-wide protection, not a per-workspace setting — Free and Teams get the same review. Full detail: Public artifacts policy.

Nothing is public by accident. Every artifact has one explicit visibility, and only you (or a collaborator with rights) can change it:

VisibilityWho can open it
privateYou, and people you explicitly share with.
workspaceMembers of the artifact’s workspace.
publicAnyone with the link — the only visibility open to the anonymous internet.

Full detail: Public artifacts policy → Visibility levels.

Personal (free) workspaces cap membership at 3 members. Teams workspaces are unlimited. See Teams overview → Seats.

Workspace data-connector credentials (API keys, OAuth tokens) are encrypted at rest with AES-256-GCM, proxied server-side through the Worker, and never sent to the browser — the artifact calls the connector by name, and the Worker injects the credential on the outbound request. This is identical on every plan; Teams doesn’t get a stronger encryption tier, it gets more people who can manage the connection. See Workspace connections.

Platform-wide, applied per account or per IP:

ActionLimit
Publish30 / hour
AI chat100 / hour
Anonymous requests100 / minute per IP