Acceptable Use & Content Policy

DRAFT — not reviewed by counsel. The Terms of Service control if there is ever a conflict.

This is the plain-language version of what you can and can't post on Talkshi.

The short version

Talkshi exists so people can talk shit with receipts. Reviews should describe what actually happened to you. Community-listed agent profiles may summarize factual public information from their linked source.

You CAN

• Name real companies and real people — public figures and private individuals — by full name.
• Criticize them harshly. Negative opinions, strong language, one-star verdicts, and "this person/company was bad at X" are all fair game.
• Make unflattering factual claims about what you experienced, and accuse someone of doing something wrong, as long as it's your own real, first-hand experience and you're not knowingly lying.
• Link to receipts — a webpage, repo, file, PR, or gist that backs up your review.
• Launch a public AI agent profile with accurate metadata and a specific public source URL. A Community-listed profile does not claim that the submitter owns, operates, or has used the agent.

We do not remove a review just because it's mean, names someone, accuses someone of wrongdoing, or makes them look bad. That's the product.

You CANNOT

These lines exist to stop genuinely dangerous or illegal content, not to protect anyone from criticism.

1. Private personal data (doxxing). No home/physical address, personal phone number, personal or someone-else's email, government IDs (SSN, passport, driver's license), bank-account or card numbers, precise location, private medical or sexual details, or passwords/secrets. Naming a person and describing their public conduct is fine; publishing their private contact info or identifiers is not.
2. Threats and incitement. No credible threats of violence, no urging others to hurt or harass a specific person, no "someone should go to their house."
3. Sexual content involving minors — never, in any form.
4. Non-consensual intimate imagery. No nude or sexual images of a person shared without their consent.
5. Illegal content of any kind, or content that helps someone break the law.
6. Spam and junk. No ads, SEO link-stuffing, fake reviews, fake profiles, duplicate launch floods, promotional profile copy, or bot floods.

Write it like a lawyer will read it

You — not Talkshi — are responsible for what you post. Keep agent profile metadata factual and supported by its linked public source. Habits that protect you and keep your review up:

• Stick to what you personally experienced.
• Say "in my opinion" when it's an opinion, and keep factual claims accurate.
• Don't state as fact something you only suspect.

How we enforce this

• Every review goes through a verified-email check and an automated safety screen before it publishes. Public agent launches are validated, safety-screened, and rate-limited by source.
• We can remove, label, or refuse any content that breaks this policy, at our discretion.
• Repeatedly breaking the rules can get your account suspended or terminated.

Reporting

If a profile or review violates this policy — especially doxxing, a threat, NCII, or content involving a minor — or if you're the subject and want it removed or corrected, use the report form (details in Report Abuse / Notice & Takedown). Urgent safety issues: the report form or [email protected].