Privacy Policy

1. What information we collect

Information you provide or direct us to retrieve includes:

• Account details: name, email, profile/preferences and sign-in information handled through our authentication provider.
• Billing information: payment and subscription details processed through Stripe. We do not store full card numbers.
• Marketing content: brand/profile and listing details, captions, uploaded media, designs, imported Canva exports, scheduled posts and other content you create, upload or ask us to retrieve.
• Social connections: account, Page, board or location identifiers, permissions, access/refresh tokens where supplied, publishing references and available insights used by the feature. We use supported authorization flows rather than asking for your social-platform password.
• Canva connections: Canva profile identifier and display name, permissions, OAuth access/refresh tokens, design identifiers/metadata and selected exported files. We encrypt those tokens at rest. We do not receive or store your Canva password.
• Optional-feature inputs: prompts, selected images, approved brand context, public website URLs and market-report location information needed for the chosen feature.
• Support messages: information you send us when requesting help.

Use of the Service generates account/connection identifiers, page views, selected actions, publishing/subscription events, counts, durations and diagnostic/security records. Requests can also carry device, browser, operating system, IP address and similar log information. The records processed depend on the feature, application controls and configured service.

2. How we use your information

We use information to manage accounts; organize, customize, schedule and publish content at your direction; connect selected social/design tools; process payments and subscriptions; deliver requested features; provide support; understand selected usage patterns; diagnose failures; prevent abuse and unauthorized access; and meet applicable legal, tax, accounting and security obligations.

Resend delivers account and service messages, including welcome, trial, payment, billing and account notices. These are separate from the weekly marketing-tip and tutorial digest.

The weekly-digest process selects accounts whose stored digest preference does not mark them opted out. This is an opt-out-based selection process, not a statement that every recipient separately signed up for marketing. A digest's signed unsubscribe link saves an opt-out preference used for future digest selection. It does not turn off separate account/service notices or recall messages already delivered. Contact support@agentpostkit.com for help.

The app does not provide a profile-settings control for re-subscribing to the digest.

3. When we share information

We share information needed for these functions. Connected third-party services also apply their own terms, controls and privacy policies.

• Supabase: account authentication, application database and media storage, including content and connection records.
• Vercel: hosting and execution of pages, APIs and static assets, including hosting request/log processing separate from browser analytics choices.
• Stripe: payment, subscription, Checkout, customer-portal and billing-event operations and their account/customer references and transaction details.
• Resend: recipient details and rendered service or digest messages, including applicable unsubscribe links.
• PostHog: product measurement such as page views and selected feature, account and subscription events, subject to Section 4.
• Sentry: error diagnostics and bounded operational metadata, subject to configured restrictions and scrubbing.
• Upstash/Redis: request-rate limits and related operational state, including hashed limiter identifiers and allow/block counts when rate-limit analytics is enabled. A hash is not necessarily anonymous information.
• Canva: authorization, connection, design-selection and export requests for the actions you choose. Imported exports can be stored in your media library.
• Connected social platforms: content, account references, permissions and operation data needed for supported actions you request through a connected platform. Availability depends on the feature, plan and provider access.
• AI services: DeepSeek or OpenAI as applicable to the feature and configured processing path, including configured fallback, as explained in Section 9.
• RSS feeds: When you add or preview a feed, or when a configured feed is polled, our servers contact its feed origin to retrieve content. We store the URLs and settings of feeds you save; polling can also store item identifiers, titles, links, excerpts, image references and publication dates.
• Optional browser push notifications: If you subscribe to browser push notifications, we store a device-specific subscription endpoint and the cryptographic material needed for delivery. We send encrypted notification payloads, which can include lead and publishing updates, to the browser push service associated with that subscription.
• Other optional retrieval services: when the relevant feature is available and used, stock search can send queries to Unsplash/Pexels; market reports can send area, ZIP or address information to Census/geocoding services; and optional website rendering can send a submitted public URL to Browserless. External image/media links can also cause requests to their own origins.

We may disclose information when required by law, court order or governmental authority. In connection with a merger, acquisition or sale of assets, your data may be transferred. We will notify you before your data is transferred and becomes subject to a different privacy policy.

We do not sell your personal information.

4. Cookies, analytics, monitoring and operational controls

Cookies and similar technologies support sign-in and preferences. Browser settings can refuse or remove storage, but doing so may affect sign-in or other functions. Authentication storage should not be assumed to expire simply when the browser closes.

PostHog product measurement

PostHog measures page views and selected actions without automatic interaction capture or session recording. Signed-in dashboard activity is identified by an internal user ID rather than your email address. Internal IDs and event references can still relate to an account. Before sign-in, signup activity can be measured using an anonymous browser identity and sign-in method category.

Browser measurement honors Do Not Track set to 1 and applicable stored opt-out choices. There is currently no public account-settings control for changing your account analytics preference. A browser preference should not be assumed to apply to every device or to activity measured by our servers.

Activity measured by our servers has separate preference and request controls. Billing-event measurement uses an internal user ID and bounded subscription/payment-event properties; these billing events are sent only when your account preference is confirmed to permit analytics. If that preference cannot be confirmed, those billing events are not sent. Browser Do Not Track settings are not automatically available to payment notifications or background work, so the same controls do not necessarily apply to every server event.

These browser choices do not by themselves stop hosting logs, billing records, security diagnostics or rate limiting. Contact support@agentpostkit.com about privacy requests.

Diagnostics and operational state

Sentry helps diagnose errors. Our collection settings restrict categories such as user information, cookies, request bodies and AI inputs/outputs, and scrub diagnostic records. Browser tracing and session replay are disabled; server diagnostics have separate settings. These controls are not a guarantee that no personal information can ever enter a diagnostic event.

Upstash/Redis supports request-rate limits using hashed request identifiers and can record allow/block counts. These operational records are separate from browser product analytics. A rate-limit window does not establish how long analytics, diagnostics or other Redis records are retained.

5. How we protect your data

We use measures such as encrypted transport, access controls and encryption of stored social/Canva connection tokens. Supabase provides the account, database and storage infrastructure described above. No internet service can guarantee perfect security. If you believe your account has been compromised, contact support@agentpostkit.com.

6. Retention, disconnecting accounts and deletion

We keep personal data while your account is active or while it is needed to provide the Service, subject to the purposes and exceptions below. See our account and data deletion instructions.

Our deletion process is designed to remove AgentPostKit-controlled product data and eligible user-owned media, verify required cleanup, then close the account. Submitting a request is not confirmation that every step has finished. Cleanup can require retry/support handling; retain the supplied request reference if you need help.

Removing a local social connection removes its stored local connection data. External grant revocation depends on the platform, credentials and implemented flow; some flows remove local credentials without revoking the external grant. Use the provider's account controls if you also want to remove its access, subject to that provider's process.

Disconnecting Canva removes the stored Canva connection and its OAuth tokens. Images you already imported from Canva remain in your private AgentPostKit media library until you delete them or complete account deletion. Disconnecting does not erase the original Canva design or other copies held by Canva.

Deleting AgentPostKit content or an account does not itself remove published social posts or copies held by platforms or their users. Backups, required records and processor logs/events have separate lifecycles. Resetting a browser analytics identity is not erasing the provider's event history.

We may also retain billing, tax, accounting, and security records where required by law or needed to protect the Service; for example, billing records may be retained for up to 7 years.

Limited operational records can remain to document request status, completion, retries and security. An identifier or digest does not necessarily make a record non-identifying or impossible to link to an account.

Ownership and the limited content license remain governed by the existing Terms of Service. This policy does not expand that license or change subscription, trial, renewal, cancellation or refund terms.

7. Your privacy choices and rights

Depending on where you live and which rules apply, you may be able to request a copy of your personal data; correction; deletion; restriction or objection to processing; a machine-readable copy; or withdrawal of consent where processing depends on consent. Requests can be subject to applicable exceptions and proportionate identity checks.

To make a request, email support@agentpostkit.com. We will respond within 30 days. Digest unsubscribe and the browser controls above are separate choices with the limited effects described in those sections.

8. Children's privacy

AgentPostKit is not for people under 18. We do not knowingly collect personal information from children. If you believe a child has provided personal information to us, contact us so we can review and delete it where required.

9. Optional AI features and content-production tools

Some tools draft or rewrite marketing content. When you choose an available AI feature, DeepSeek or OpenAI can receive the request through its configured processing path. Depending on the feature, inputs can include entered text, selected images, instructions, approved brand/profile context or retrieved facts—not only the words entered into one form. A configured fallback can cause an alternative provider to receive the request.

Market reports can combine location-derived Census facts with approved brand context and save report inputs, derived data and generated captions/hashtags. Notices and controls are feature-specific; market-report generation asks for your consent and checks your AI Studio consent. This is not a promise that every AI feature uses an identical one-request consent mechanism. You can skip optional AI tools and create content manually where that workflow is available.

An administrator-only draft tool can send brand-voice instructions, category/title hints and a template-thumbnail URL as an image input to OpenAI to prepare library metadata/captions. Where staff use operator tools, OpenAI or ElevenLabs can process provided text for content or narration. Saved audio can later be served as a static asset without another generation request. These separate flows do not mean every customer's data is sent to operator tools or that admin access establishes customer consent or image rights.

Generated content and related records can be saved for the feature's use. Provider treatment depends on the service, API, agreement and settings. This policy makes no zero-retention guarantee. The Terms' limited content license, including its limits on general-purpose AI-model training, remains unchanged; provider statements do not expand it. Do not put payment details, passwords, connection tokens or unnecessary sensitive information into submitted content.

10. Policy updates

We may update this Privacy Policy from time to time. We will notify you of material changes by posting the new policy on this page and updating the "Last updated" date, and where appropriate, by sending you an email notification.

11. Contact us

Questions about this policy or how we handle data? Contact AgentPostKit at support@agentpostkit.com.