Editorial promise
Plain language, operational consequences.
We do not treat fluency as autonomy or a model as a whole service. Every guide starts with a user goal, describes system behavior, names failure, and links to inspectable sources.
About this field guide
For people who have to make the conversation work.
Chatbot Field Guide is a proposed practical publication for product teams, builders, service operators, and users. It explains the interaction, names the failure mode, and makes the operator’s responsibility visible.
What stays out
No agent magic.
No deceptive humanlike design, unverified autonomy claims, invented people, synthetic interviews, imaginary tests, or backdated publication history. Product documentation is useful evidence of documented behavior, not independent proof of quality.
Current status
A private review build.
The profile remains a research candidate. Its domain is null, the manager is unassigned, the newsletter is unconfigured, and the source imagery awaits owner publication review.
Corrections belong in the record.
This preview has no working contact or submission backend. During review, corrections should be attached to the local source package and coverage ledger so claims, dates, and article mappings can be revised together.