Every visit gets two answers: a verdict (how sure Loghound is that a person was driving) and a class (what kind of automation it was, if any). Here are both, with the identifier each one is stored as.
01 · Verdicts
| Identifier | Label | Meaning |
|---|---|---|
human | Human | Scored low enough that we believe a person drove this visit. |
likely_human | Likely human | Some weak automation signals, not enough to call it a bot. |
unknown | Unknown | Scored, and the evidence did not reach a verdict either way. Not a failure to measure. |
likely_bot | Likely bot | Enough signals to suspect automation, short of the bot threshold. |
bot | Bot | Automation, honest or not. Look at the class to tell which. |
The score bands behind them and the checks that move a verdict: How a visit is scored.
02 · Bot classes
| Identifier | Label | Meaning |
|---|---|---|
declared_crawler | Declared crawler | Said what it was in the User-Agent and was telling the truth. Honest traffic. |
ai_crawler | AI crawler | A declared crawler collecting for model training or for answering in an assistant. |
monitor | Monitor | A declared uptime or availability checker. |
spoofed_ua | Spoofed User-Agent | The headers contradict each other, or a crawler claim failed reverse DNS. |
proxy_fleet | Proxy fleet | One browser fingerprint arriving from many unrelated networks. |
headless | Headless browser | A driven or screenless browser: an automation marker, a software renderer, or a failed engine claim. |
undeclared_client | Undeclared client | Not a browser, and did not declare itself a known crawler: software that did not say what it is. |
scripted | Scripted client | Scored as automation without fitting a named class, including a declared client that probed for attacks. |
none | Not automation | Not classed as a bot of any kind. |