Acceptable Use Policy

1Purpose

ProductAtlas exists to share production-informed product knowledge that people and AI agents can reuse safely. This policy keeps the catalog useful and keeps the people and agents that consume it out of harm's way.

It applies to everything you publish, upload, or transmit, and to how you use the website, the API, and MCP. Terms used here have the meanings given in the Terms and Conditions.

2Content you must not publish

• Content you do not have the right to publish, including your employer's or a client's confidential material, and content that infringes copyright, trademark, patent, or trade secret rights.
• Credentials, API keys, tokens, private certificates, or other secrets, whether yours or someone else's.
• Personal data about other people, including customer records, employee details, and screenshots containing identifiable information you have no right to share.
• Malware, exploit payloads, or instructions whose evident purpose is unauthorized access to systems or data.
• Content designed to manipulate an AI agent that retrieves it, including hidden or embedded instructions intended to override the agent's operator.
• Content that is unlawful, defamatory, harassing, hateful, or that sexualizes minors.
• Deceptive provenance: falsely attributing content to another person or organization, or misrepresenting that a Blueprint was used in production.
• Bulk low-value content, spam, keyword stuffing, or listings whose main purpose is to advertise an unrelated product.

Writing about security is welcome and expected. A Blueprint may describe attack scenarios, threat models, and defensive design. What is prohibited is content that functions as an attack tool rather than as product knowledge.

3Conduct that is not allowed

• Probing, scanning, or attempting to bypass authentication, authorization scopes, rate limits, or safety screening.
• Attempting to read another publisher's unpublished Drafts or private state.
• Automated access that degrades the service, or scraping designed to replicate the catalog wholesale.
• Creating accounts or namespaces to impersonate a person or organization, or to evade a suspension.
• Manipulating trust, usage, or quality signals, including through coordinated or automated activity.
• Misrepresenting a trust badge, for example presenting Atlas Safety Checked as a guarantee of security.
• Interfering with moderation, including republishing content that was withdrawn for a policy breach.

4Publisher responsibilities

• Review a Release before publishing it. Once published it is immutable and may be retrieved immediately by agents.
• Sanitize screenshots, logs, and configuration examples before uploading them.
• Select a license you are entitled to apply, and credit sources you built on.
• Say plainly where guidance is untested, opinionated, or specific to a context, rather than implying broader validation.
• Publish a corrected version when you learn a Release contains a material error.

5Reporting a problem

Anyone can report a Release, whether or not they have an account. Use the report action on the Release, or email [email protected].

Tell us the exact Blueprint and version, what the problem is, and any evidence you can share. Reports that identify a specific Release and category are the fastest to act on. Do not include secrets or personal data in a report beyond what is needed to explain the issue.

If you are reporting a vulnerability in ProductAtlas itself rather than in published content, email the same address and give us a reasonable opportunity to fix it before disclosing publicly.

6Copyright and rights-holder notices

If you believe content on ProductAtlas infringes your rights, send a notice to our designated copyright agent at [email protected], identifying the work, the exact Release, your contact details, and a statement that you have a good-faith belief the use is unauthorized.

Notices may also be sent to opsAI LLC, Attn: Copyright Agent, 13740 N Highway 183, Ste L2 #184, Austin, TX 78750, United States, or by telephone on +1 512-953-7797. Email reaches us fastest.

We remove infringing content from public distribution and may act against repeat infringers. If your content was removed and you believe that was wrong, you may respond and we will reconsider.

7How we enforce this policy

We act proportionately, based on the severity of the breach, whether it looks deliberate, and whether it has happened before. Serious breaches such as secrets exposure, malware, or content sexualizing minors are acted on immediately.

• Ask you to fix the problem and republish.
• Withdraw or quarantine a Release so it is no longer publicly distributed.
• Revoke a trust badge or a verification.
• Suspend or ban an account.
• Report unlawful content to the relevant authorities.

Every enforcement action is recorded as an auditable moderation event. We tell affected publishers what happened and why, and give them a route to respond, unless doing so would be unlawful or would frustrate an investigation. Actions are reversible where a report is not upheld or the issue is resolved.

8Changes to this policy

We update this policy as new patterns of misuse appear. The effective date at the top reflects the current version, and material changes are announced alongside the Terms and Conditions.