ailetterbox guidelines.

ailetterbox is a letterbox that routes AI-drafted work to a human expert for a sense-check. Seekers can be humans (via the paste box) or agents (via MCP). Experts respond with one reply with four fields — a sense-check (OK to use / don’t use), comments, an optional estimate to rework, and an optional paid-review offer. ailetterbox does not provide opinions or advice — the reply is from the independent expert, not from ailetterbox.

For experts

By registering as an expert on ailetterbox you agree to the following. Breach may lead to profile removal.

1. Truthful profile
   Your name, firm, jurisdiction, categories, skills and credentials must be true and current. Update your profile if your practice, qualifications, or licence status changes.
2. Respond within your stated window
   Every brief must be acknowledged within your stated response window (platform default: 2 working days). Accept, decline, or reply with your opinion. Silence is not an acceptable response.
3. Deliver a sense-check — not formal advice
   Your reply has four fields: a sense-check (OK to use / don’t use), comments (your reasoning), an optional estimate (a quote to rework linked to your own terms), and an optional paid review note if you’d take the brief on properly for a fee. A sense-check based on the information the seeker has provided. Not a substitute for formal advice under a signed engagement. Where a seeker needs binding advice or paid rework, engage them separately off-platform under your own terms.
4. Comply with your professional obligations
   You remain bound by whatever regulatory, licensing, and conduct rules govern your profession — for lawyers the SRA / Law Society / Bar Standards Board (UK) or State Bar (US); for accountants the ICAEW / ACCA / AICPA; for architects the ARB / AIA; and equivalents in every regulated profession. ailetterbox does not replace these obligations.
5. Run your own conflict-of-interest checks
   Before responding to any brief, run whatever conflict-check process applies to your practice. If you have a conflict, decline the brief promptly — do not review the content or reply.
6. Treat brief content as confidential
   Do not share, publish, or discuss the content of any brief you receive outside your firm, except that you may retain briefs and your reasoning in your own private Knowledge Base (visible only to you) to power your Expert AI. This is a hard rule.
7. No solicitation or marketing content
   Your reply is a professional opinion. It should not contain marketing content, cross-promotion of unrelated services, or solicitation of work in unrelated areas.
8. Decline promptly if you cannot advise
   If a brief is outside your competence, jurisdiction, availability, or contains material you cannot advise on, decline it promptly so the seeker can try another expert. Do not simply ignore it.
9. PIN hygiene
   Your PIN is a single key you share with every client you want to accept briefs from. Regenerate it whenever it may have been exposed, when a client offboards and shouldn’t keep access, or periodically as a precaution. The old PIN is invalidated the moment you regenerate — so re-share the new one with any clients you still want to hear from.
10. Same reply quality regardless of channel
    A brief routed via an AI chat (Claude, ChatGPT, any MCP-aware tool) is still sent by a human seeker — the AI chat is a formatting layer. Give these briefs the same quality of reply as directly-sent briefs. Do not downgrade the response because it arrived via MCP.
11. AI-Agents when it goes live
    If you have joined the AI-Agents waiting list and AI-Agents goes live for your profession, remember that autonomous AI requests are fee-gated and may be sent at volumes much higher than human briefs. Keep your availability and response window accurate; decline promptly if a brief is outside your competence; and apply the same conflict-check and no-advice-across-jurisdictions rules as you do for human briefs.

For seekers (humans and agents)

By sending a brief through ailetterbox — via the paste box (humans) or MCP (agents) — you agree to the following.

1. Only submit content you have the right to share
   Do not submit content that is subject to another party’s confidentiality obligations without their consent. Do not submit anyone else’s AI output without permission.
2. Understand what you’re receiving
   The reply you receive is a sense-check from an independent expert, not from ailetterbox. One reply with four fields — a sense-check (OK to use / don’t use), comments, an optional estimate to rework, and an optional paid-review offer — with the expert’s reasoning. It is not a substitute for formal advice or paid rework. For those, engage the expert directly under a formal engagement.
3. Verify credentials yourself before paid follow-on
   ailetterbox does not independently verify the credentials any expert declares. Before engaging in paid follow-on work, check the relevant professional register for that expert’s profession (regulator, chartered body, or licensing board).
4. Respect the expert’s decision
   Experts may decline briefs for many reasons — competence, conflict, time, scope. A declined brief is not a rejection of you; it’s a professional judgement. Try another expert.
5. Do not submit unnecessary sensitive information
   Avoid including highly sensitive personal data (medical records, financial account numbers, national insurance numbers) unless directly necessary to the matter.
6. No spam or abuse
   Do not send the same brief to unlimited experts to fish for opinions. Do not misrepresent the nature of a brief. Repeated blocks by experts will trigger account review.

Enforcement

ailetterbox reserves the right to remove profiles and accounts that breach these guidelines. Serious breaches (fraud, impersonation, sharing brief content externally) may be reported to the relevant professional regulator.

Written reports to WeLINKanything Ltd, 21 Church Road, Poole, Dorset BH14 8UF, United Kingdom.