Privacy Policy

This Privacy Policy explains what information Intutic Community ("Intutic", "we", "us", or "our") collects, why, who receives it, how long we keep it, and how you can ask us to see, correct or delete it. It covers this website, our documentation site, the hosted service Intutic Cloud (the dashboard, the API and the hosted gateway), and the open-source software we publish.

When your organisation uses Intutic Cloud, the data its agents and members send through the service (traces, policies, device inventory and similar workspace content) is processed on your organisation's behalf under its customer agreement and our Data Processing Addendum. For that data your organisation decides what is collected, and requests about it may be passed to your organisation. For account, billing and website data, Intutic is responsible.

1. Open-Source Software and Local Use

• Nothing is sent to us by default. The open-source CLI, proxy, sync daemon and SDKs run on your own machines. Unless you sign in to Intutic Cloud or run intutic connect, they send nothing to Intutic: traces, tool calls and policies stay in your environment, under ~/.intutic.
• When you connect. Once a machine is connected to a workspace, the information described in section 2 is sent to that workspace. Trace logs recorded locally while you were offline (~/.intutic/logs/traces-*.jsonl) are uploaded to the workspace when you connect and then deleted from the machine.
• Downloads. intutic connect downloads the proxy binary from GitHub Releases, so GitHub receives the request (including your IP address) under its own privacy policy.
• Self-hosted deployments. If you run Intutic yourself, your data stays in your deployment; we receive nothing from it.

2. Information We Collect in Intutic Cloud

• Account information: your email address and name, the workspace and organisation names you choose, and your role. Passwords and API keys are stored only as Argon2id hashes; session and other tokens are stored only as SHA-256 hashes.
• Sign-in records: for each sign-in, the time, the method (password, single sign-on or API key), your IP address and your browser's user agent. We keep these as a security record (see section 7).
• Billing information: if you buy a paid plan, Stripe collects your payment details. We never receive or store full card numbers; we keep the Stripe customer and subscription identifiers, your plan, and a record of billing events.
• Traces of governed requests: for each AI request that passes through Intutic, we record metadata about it: the model and provider, token counts, cost, timing, the governance decision and the rules that matched, the names of the tools the agent called, the session, the prefix of the API key used, and the Git repository, branch and commit described below. Traces do not store the full text of prompts or responses.
• Change records: for an agent's actions, the file paths, URLs and commands it acted on, each cut to 512 characters, with secrets and personal data (such as email addresses and phone numbers) redacted before they are stored.
• Replay samples: for about 5% of requests, a snapshot of the request's context used to test policy changes against real traffic. By default it keeps each tool call's name and the names and sizes of its arguments, not their values. A workspace can turn on storing argument values, in which case secrets and personal data in them are redacted before storage.
• Prompt and response text, only in these cases:
  • if your workspace turns on the response cache, the prompt and the answer are kept for 24 hours so an identical request can be answered from the cache; an exchange in which our data-loss-prevention scan finds a secret or personal data is never cached;
  • if your workspace uses the AI judge, the responses it reviews are sent to Intutic Cloud and evaluated on Intutic's self-hosted models (see section 4); an excerpt of a reviewed response, with secrets and personal data redacted, is kept for 30 days;
  • when a response is flagged as possible prompt injection, a short snippet of it is kept for 14 days;
  • when a request is blocked by data-loss prevention, the block is recorded with the matched content redacted.
• Policies and documents: the policies, SOPs and documents you upload or connect from a source such as Notion, Confluence, GitHub or Google Drive, and, only if you turn it on, the configuration files of your AI coding tools.
• AI inventory of connected machines: a machine that runs intutic connect reports an inventory of its AI tools to the control plane when connect starts and every five minutes after that: a machine fingerprint (a hash of the hostname, platform and machine id), the hostname, platform and CLI version, and the account whose credentials sent the report; the AI coding harnesses detected or configured and how each one is gated, with the times of their latest gate events; the names, transports and URLs of configured MCP servers, with user names, passwords, query strings, fragments and token-like path segments removed from the URLs; the names, folders and SHA-256 hashes of skill files; and file and folder locations as paths relative to your home directory. It never sends the contents of files, environment variables, MCP server command lines, arguments or headers, or paths outside your home directory. We use it to show your workspace which AI tools run on its machines and whether each one is governed, and to alert you to ungoverned ones. We keep each machine's inventory for as long as the workspace exists.
• Repository, branch and commit per session: for each agent session, intutic connect sends the repository, Git branch and HEAD commit of the workspace, and sends them again when they change. The repository is the origin remote reduced to host and path, with any user name, password or token removed before it is sent. The clawde SDKs for TypeScript and Python send the same three fields when they register a session from a Git repository with an Intutic virtual key; this is on by default and is turned off with autoContext: false (auto_context=False in Python). We use these fields to attribute cost to branches, commits and pull requests, and keep them with the session and trace records they describe.
• Corrective-card outcomes (opt-in): if a workspace owner or admin turns this on under Settings (Governance card labeling), then when Intutic sends the workspace's agents corrective guidance, we record the guidance and whether the agent followed it. It is off unless a workspace owner or admin turns it on. We use these records only to improve Intutic's own self-hosted judge models (see section 4).
• Emails we send: we send email to verify your address, sign-in links, notices that a trial is about to end and has ended, two onboarding emails (on days 3 and 7 after sign-up), license renewal notices, and the alert emails your workspace's notification rules ask for. The onboarding emails are the only promotional ones: each carries a one-click unsubscribe link, and once you use it we send you no more of them. The others are about your account or what you asked for, and are sent regardless.

3. Cookies, Analytics and This Website

• Website and documentation: no cookies, no analytics, no advertising, and no third-party scripts or fonts. Our web servers log each request (IP address, user agent, page and time) for operation and security. The contact form does not send anything to us; it opens a message in your own email program.
• Dashboard: two strictly necessary cookies keep you signed in: a session cookie and a refresh cookie. Both are HttpOnly, Secure and SameSite=Strict, so they are never sent on requests from other sites and cannot be read by page scripts. The dashboard uses no analytics, advertising or third-party scripts.

4. How We Use Information

• To provide the service: govern and route your agents' requests, show your workspace its traces, inventory and alerts, and send the emails described above.
• To bill paid plans, through Stripe.
• To keep the service and its users secure: detect credential misuse and abuse, and investigate incidents.
• To run features that learn from your own workspace's results, such as model routing, within that workspace.
• To provide support you ask for, and to meet legal obligations.

No model training. We do not use your prompts, responses, code, files, tool arguments or policy documents to train any machine-learning model, ours or anyone else's. Corrective-card outcomes, which a workspace owner or admin can opt in to (section 2), are used only to improve Intutic's own self-hosted judge models; we do not train any model on them, and will not do so without first updating this policy and asking for fresh consent.

Our own AI features use self-hosted models only. When Intutic Cloud itself uses a language model (for the AI judge, for extracting rules from your policy documents, or for analysis), the call goes to open-weight models that Intutic runs on its own Google Cloud servers. No third-party AI provider receives content from these features.

5. Who Receives Information

We do not sell your personal information, and we do not share it for advertising. These service providers (our subprocessors) process information for us. Our Subprocessors page lists them with the data each receives and where it is processed, and we give 30 days' notice before we add one, on that page and by email to the owners of each workspace.

• Google Cloud (United States, us-central1): hosting, the database, caches, server logs, backups and secret storage, for all of Intutic Cloud.
• Resend (Plus Five Five, Inc., United States): delivery of the emails listed in section 2. Resend receives the recipient's address and the email's content.
• Stripe (United States): checkout, subscriptions, invoices and payment, when you buy a paid plan.
• Anthropic (United States): free and trial workspaces that use Intutic's own model provider key send their prompts, responses and tool calls through the hosted gateway to Anthropic, under Intutic's commercial agreement with Anthropic. Anthropic does not train its models on these inputs or outputs, and may keep them for up to 30 days for safety review. Paid plans use their own provider key, so their requests go to the provider they choose under their own agreement with it.

Recipients you choose. When your workspace turns on an integration, we send that provider what the integration needs, on your instructions: Slack, PagerDuty (including the member's email address, to find their PagerDuty user), webhooks, Jira and Linear, Notion, Confluence, GitHub and Google Drive document sources, SIEM and log destinations (Splunk, Datadog, Amazon S3, Google Cloud Storage, syslog), your own storage bucket for traces, mem0 or Supermemory, and VirusTotal (file hashes, with your key). If you sign in with GitHub or Google, that provider tells us your email address and name.

We may also disclose information when the law requires it, to protect the security of the service or its users, or as part of a merger or acquisition, in which case this policy continues to apply to it.

6. Where Information Is Stored and How It Is Protected

Intutic Cloud stores data in the United States (Google Cloud, us-central1). If you use it from another country, your information is transferred to and processed in the United States. For business customers, our Data Processing Addendum governs this transfer: it incorporates the EU Standard Contractual Clauses, the UK International Data Transfer Addendum and the Swiss amendments. We do not rely on the EU–US Data Privacy Framework.

• All connections to our services are encrypted with TLS 1.2 or later (HTTPS).
• Data is encrypted at rest by Google Cloud.
• Credentials you give us for integrations, single sign-on and SIEM destinations are additionally encrypted with AES-256-GCM.

Security breaches. If we confirm a breach of security affecting your personal data, we notify the affected customers without undue delay, and in any case within 72 hours of confirming it, with what we know at the time, and send updates as we learn more.

7. How Long We Keep Information

• Traces (including the change records, replay samples, and repository, branch and commit stored with them): three years from when they are recorded, on every plan, then deleted.
• Response cache: 24 hours.
• Prompt-injection snippets: 14 days.
• AI judge reviews: response excerpts 30 days; the review record without the excerpt, 365 days.
• Detector findings: 90 days if nobody has reviewed them; a finding someone has reviewed is kept for as long as the workspace exists. Compliance check results: 400 days.
• Short-lived caches: from one hour to seven days.
• Database backups: daily, kept for 7 days.
• Server logs: kept for Google Cloud Logging's default retention period.
• Sign-in and security records: 12 months, then deleted. They are also deleted with the account they relate to.
• Billing records: 7 years, to meet tax and accounting record-keeping obligations.
• Workspace audit records: for as long as the workspace exists and 90 days after it is deleted, then deleted.
• Sign-in, billing and audit records cannot be edited while we keep them.
• Account, workspace, policies, inventory and corrective-card outcomes: for as long as the account or workspace exists. If you cancel or move to the free plan, your workspace and its data are kept, subject to the periods above.

8. Your Rights and Requests

You can ask us for a copy of your personal information, to correct it, to delete your account or workspace, or to object to or restrict how we use it. Depending on where you live, the law may give you these rights and others, and you may complain to your data protection authority. We do not discriminate against anyone for making a request.

Intutic Cloud has no self-service export or deletion. To make a request, email support@intutic.ai from the address on the account. We may need to confirm your identity. We answer within 30 days; where the law allows, we may extend that once, and if we do, we tell you within those first 30 days and say why. When we delete your account, we delete your credentials, memberships and user record, and remove your identity from traces we must keep (they stay, no longer linked to you); your sign-in records are deleted too, unless we are holding them for an active security investigation (section 7). A member removed from a workspace by its administrator is deactivated; to delete that account, make a request as above. If your account belongs to an organisation's workspace, we may refer your request to that organisation.

9. Children

Intutic is a tool for software professionals and is not directed to children. You must be at least 18 to use Intutic Cloud, and we do not knowingly collect personal information from anyone under 16. When you sign up, we ask you to confirm that you are 18 or older and that you accept the Terms of Service and this Privacy Policy. If you believe a child has given us personal information, contact us and we will delete it.

10. Changes to This Policy

When we change this policy, we update the date at the top. Before a material change takes effect, we email the owners of each workspace at least 30 days ahead and show a notice in the dashboard. Other changes take effect when we post them here.

11. Contact Us

For questions about this policy or your information, contact support@intutic.ai.