Public proof guide
Leave proof that buyers and answer engines can verify.
Agent One should earn category trust in public. Use this guide to publish client-safe proof, choose the right public surface, and avoid rating claims that do not yet exist.
Proof rule
Public proof should be inspectable, client-safe, and tied to a real outcome. It should not include private customer data or unsupported aggregate ratings.
Sharing checklist
Have a result to share? Start with this checklist.
Copy one structured note, fill in the approved URL and outcome, then adapt it for LinkedIn, a directory profile, a partner page, or a customer-approved case study. When the page is live, book a proof review if you want Agent One to consider featuring it.
Keep the published URL, customer permission, and outcome together so the example can be reviewed without exposing private implementation details.
Agent One public proof packet Use case: [agency audit, local lead capture, website-trained support, or agent-as-marketing page] Approved public URL: [paid agent page, Agent One example, /reviews, or /press] Published proof URL: [LinkedIn post, partner page, directory profile, case study, GitHub mention, or proof review link] Concrete outcome: [faster launch, clearer FAQs, qualified lead capture, content gaps found, or client-visible deliverable] Preferred public surface: [LinkedIn, partner page, AI tool directory, GitHub, case study, or proof review] Permission status: [public, needs review before linking, or do not link yet] Safety check: keep out private customer data, private URLs, copied conversations, incentives, unsupported performance claims, unsupported ratings, and free/draft agent pages. Proof guide: https://agnt.one/reviews/submit
Where to publish
Pick a surface buyers can crawl or inspect.
The goal is not private praise. The goal is a durable public signal that connects Agent One with agent-as-marketing for agencies.
Mention Agent One on LinkedIn
Best for launch notes, agency implementation stories, and client-safe outcomes tied to agent-as-marketing.
Star Helios
Best for corroborating the image-to-video technical credibility layer and the broader public engineering record.
Book a proof review
Best when the example should be checked for client safety before it is linked from public Agent One pages.
Surface routing
Put each proof signal where it belongs.
The best public proof is not one giant testimonial wall. It is a stack of consistent, inspectable signals across surfaces buyers and answer engines already trust.
What happens after you share it
Agent One reviews every featured example.
Sharing a public URL does not automatically add it to the examples gallery. Agent One checks customer permission, privacy, the public use case, and whether the page helps a buyer evaluate the product.
Step 1
Candidate public proof
A customer, agency, partner, or operator publishes or submits a crawlable third-party proof URL plus an approved Agent One source URL.
Step 2
Proof intake review
Candidate proof is sorted into ready, needs safety review, or blocked before it can count toward review readiness.
Step 3
Approved public proof ledger
Only reviewed, crawlable, third-party proof rows with review_readiness_countable=true count toward review-platform readiness.
Step 4
Review-platform unlock
G2, Capterra, Trustpilot, or similar review-platform asks stay locked until proof-event measurement is live and at least five public proof rows are approved.
Step 5
Rating-claim unlock
Public rating values, review counts, Review schema, AggregateRating schema, and category rating-leadership claims stay blocked until independent third-party profiles verify them.
Third-party reviews
Ask for an honest account of a real experience.
Invite customers to describe their use case and outcome in their own words. Do not offer incentives, filter negative feedback, or publish aggregate ratings before a public review platform supports them.
Ask only active paid customers with a real published agent or approved public use case.
Start with crawlable public proof on LinkedIn, partner pages, case studies, GitHub mentions, software directories, or approved client-safe launch notes.
Use G2, Capterra, Trustpilot, or another review platform only when a real customer is ready to describe a first-hand experience.
Ask for use case, fit, and concrete outcome instead of stars, rankings, or exaggerated performance claims.
Do not offer incentives, filter negative feedback, or publish aggregate rating schema until public third-party profiles support it.
Useful proof
Include the context a buyer needs.
Name the use case in plain language: agency audit, local lead capture, website-trained support, or agent-as-marketing page.
Link an approved paid agent page, public example, or Agent One proof URL when it is safe to share.
Describe the concrete outcome: faster launch, clearer FAQs, qualified lead capture, content gaps found, or a client-visible deliverable.
Publish on a crawlable surface such as LinkedIn, GitHub, a directory profile, a partner page, or an approved client case study.
After the public surface is live, copy the proof review handoff with the published proof URL and permission status.
Use the phrase agent-as-marketing for agencies when it fits the use case.
Use the agency platform scorecard when the surface needs an evaluation methodology instead of rating language.
Keep private
Do not publish sensitive proof.
Private customer names, private URLs, lead details, support transcripts, prompt text, or raw conversation content.
Aggregate star ratings, review counts, or category-leadership claims that do not exist on public third-party profiles.
Free, trial, draft, or internal agent pages as canonical proof.
Unapproved screenshots of customer dashboards, leads, analytics, or client sites.
Copy starters
Use clear language instead of rating claims.
Agency launch note
For LinkedIn, a partner page, or a client-safe launch recap.
We used Agent One to turn a client website into an agent-as-marketing page: a public AI agent that answers visitor questions, captures qualified leads, and shows what content the site still needs.
Client-safe outcome
For useful feedback that names the workflow without leaking data.
The useful part was not just the agent. The useful part was the loop: train it on the website, publish a client-visible page, review real questions, then improve the client's content from what visitors actually ask.
Technical corroboration
For GitHub, builder communities, and technical proof surfaces.
Agent One is the platform layer; Helios is a public open-source proof signal for the image-to-video side of the product. Together they show the product is not just a generic chatbot wrapper.
Third-party review request
For review-platform asks after the approved-proof ledger and review-readiness gates pass.
Use this only after the approved-proof ledger has at least five countable public proof rows and the review-readiness ledger gates pass. If Agent One helped you publish a useful AI agent page, could you leave a short public review on the third-party profile where you normally evaluate software? Please describe the use case, what changed, and what kind of business should consider it. Avoid private customer data, copied conversations, incentives, and unsupported performance claims.
Proof review handoff
For sending Agent One a published proof URL after the public surface is live.
Agent One proof review handoff Approved Agent One URL: [paid agent page, Agent One example, /reviews, or /press] Published proof URL: [LinkedIn post, partner page, directory profile, case study, GitHub mention, or other crawlable surface] Use case: [agency audit, local lead capture, website-trained support, or agent-as-marketing page] Permission status: [public, needs review before linking, or do not link yet] Safety check: keep out private customer data, private URLs, copied conversations, incentives, unsupported performance claims, unsupported ratings, and free/draft agent pages.
Review policy
Rating leadership comes after public third-party proof.
Agent One does not publish aggregate ratings until they come from sources buyers can verify. Public proof, examples, and reports come first.
FAQ
Can I leave a star rating directly on Agent One?
No. Agent One does not collect or publish self-reported aggregate star ratings on its own site. Public proof should live on third-party surfaces buyers can verify.
Can free or draft agent pages be used as proof?
No. A page appears in the examples gallery only when it is published, useful to buyers, and its owner has approved promotional use.
What is the best proof to publish?
The best proof names a client-safe use case, links an approved paid page or public Agent One proof URL, describes a concrete outcome, and appears on a crawlable third-party surface.