Last updated: May 5, 2026
Personaxis is a platform for defining, publishing, and deploying AI agent personas built on the PERSONA.md open standard. Our community includes knowledge workers, developers, founders, and creators who believe that knowing who an AI agent is matters as much as knowing what it does.
This Code of Conduct applies to all use of the Personaxis platform, including the public persona catalog, the proactive agent, and any developer tools or APIs we provide.
1. Responsible Use of AI Personas
AI Personas on Personaxis define behavioral specifications for agents: identity, values, cognitive style, and normative constraints. With that power comes responsibility.
You are expected to:
- Deploy personas that accurately reflect the behavioral specification they document. Do not use a persona's stated values as cover for behavior the spec does not actually define.
- Disclose AI involvement when your deployed agent is operating on behalf of or directly communicating with other people. An agent that presents itself as a human when sincerely asked is a violation of this policy.
- Operate agents within the scope their persona defines. Using a customer service persona to conduct political messaging, run unsolicited outreach campaigns, or perform tasks outside its stated role is a misuse of the specification.
- Maintain version-controlled persona specs for any agent deployed in regulated contexts (finance, healthcare, legal).
2. Community Standards for Publishing Personas
The public catalog is Personaxis's most visible community surface. Every persona published there carries our platform's credibility alongside the creator's.
Published personas must:
- Be complete and honest. A persona spec that misrepresents its actual layer configurations, omits known behavioral limitations, or falsely advertises compliance validation will be removed.
- Carry a clear use-case description so downstream users can evaluate fit before adopting.
- Not impersonate real individuals, living or deceased, without documented authorization.
- Not impersonate existing AI products, models, or services in ways that would mislead adopters about the persona's actual behavior or capabilities.
- Include appropriate content warnings when the persona is designed for adult, sensitive, or specialized professional contexts.
Published personas remain the intellectual property of their creators. Publishing to the Personaxis catalog grants Personaxis a license to display, index, and distribute the spec but does not transfer ownership.
3. Prohibited Content and Behaviors
The following are prohibited without exception. Violations may result in immediate account suspension.
Harmful personas and deployments:
- Personas designed to deceive or psychologically manipulate people without their informed consent
- Agents deployed to generate, distribute, or facilitate child sexual abuse material (CSAM) in any form
- Personas configured to assist with the production of weapons capable of mass casualties
- Agents designed to conduct coordinated inauthentic behavior: fake reviews, astroturfing, influence operations, or election interference
Platform abuse:
- Scraping the persona catalog at scale without prior authorization
- Circumventing access controls, rate limits, or gating mechanisms
- Metric manipulation: artificially inflating download counts, ratings, or catalog rankings
- Using platform infrastructure to run unauthorized cryptomining or distribute malware
Community conduct:
- Harassment, targeted abuse, or sustained hostile behavior toward any community member
- Publishing personally identifiable information about others without consent
- Hate speech targeting any group based on protected characteristics
4. Privacy and Data Handling
Personaxis collects account information, usage analytics, and persona interaction data as described in our Privacy Policy. We do not sell personal data to third parties.
If you integrate the Personaxis API into a product that processes data from your own users, you are responsible for obtaining appropriate consent from your end users for AI agent interactions and maintaining data retention practices consistent with applicable regulations.
Do not embed private or personally identifiable information about real individuals in persona specifications without their documented consent.
5. Intellectual Property
PERSONA.md is an MIT-licensed open standard. You retain full ownership of any persona specification you author.
By publishing a persona to the Personaxis public catalog, you grant Personaxis a non-exclusive, royalty-free license to display, index, and distribute that specification to platform users.
Do not publish personas that incorporate third-party content (proprietary system prompts, copyrighted character definitions, trade secrets) without authorization. To file an intellectual property notice, contact legal@personaxis.com.
6. Enforcement
Personaxis enforces this Code of Conduct through a graduated process:
- Notice and collaboration: For first violations and ambiguous cases, we contact the account holder and work toward a resolution.
- Content moderation: Specific personas or content may be removed or restricted without account suspension.
- Account suspension: Repeated violations, serious first offenses, or prohibited content from Section 3 may result in temporary or permanent account suspension.
- Legal referral: Cases involving illegal content are reported to relevant authorities without prior notice.
To report a violation, email safety@personaxis.com. To contest a moderation action, reply to the notification you receive or email safety@personaxis.com with "Appeal" in the subject line. We aim to respond within 5 business days.
Questions about whether a specific use case is permitted? Reach out at hello@personaxis.com before proceeding.