How WeBot prepares public guides and examples
Public information should help you choose a task and understand the inputs before using credits. This policy explains the evidence labels used by the workflow resources published with this update.
Product and workflow facts
Workflow names and route links are checked against the application’s public model registry. Availability and selectable settings remain controlled by the live workspace. Guides do not override account, privacy, usage or payment terms. Older blog content and external model documentation may need their own review when a provider changes a feature.
Examples and provenance
An output example must identify its actual public sample and the input context available for it. We do not label a prompt recipe as a generated result. If the source does not supply a seed, run date or cost, the example does not invent one. Curated samples are examples, not a measured success rate.
Comparisons and commercial claims
A task comparison may explain supported input types without a fresh generation test. Such pages disclose that limit. A hands-on quality, speed or cost comparison needs actual results, test conditions and evidence. We do not publish fabricated benchmark scores, customer quotes, usage statistics or case-study outcomes.
Customer material
A customer case study needs the customer’s permission and verified project material before publication. Until those are available, a proposed scenario stays a planning guide. Model and company names identify the relevant products and do not imply an endorsement or partnership.
Corrections and review
The new workflow resources state the review context where relevant. If a fact becomes unclear, send its page URL and the supporting source to support@webot.online so it can be reviewed. A date is not evidence that a model’s output quality or every older article was retested.