Read the dog in front of you.

Record what you can actually see and hear, and the situation you are in. You will get the most consistent reading, the competing one, how confident that is, what to do next, and the full reasoning behind it.

This does not translate barks. There is no microphone here and nothing is recorded. It reads the pattern you describe and tells you what that pattern is associated with, with the evidence named. What it can and cannot do.

Reading for

0 signals noted

Why it will not answer from one signal. A single cue is the most common way a dog gets misread, so the interpreter requires the context plus at least two body signals before it will offer a reading. This is a hard rule in the code, not a suggestion. See the three locks.

Reference

What this interpreter reads

The tool above is a transparent scoring table, not a mind and not a microphone. These are the exact states it weighs and the confidence bands it can reach, the same set the receipt shows for any reading. It never turns a sound into words, because nothing can.

The ten states it weighs

Every reading is scored across this one fixed set. A dog is never handed a sentence: you get the most consistent state, the competing ones, and the evidence behind each.

Relaxed / affiliative

Loose, settled, and comfortable in company.

Playful

Actively inviting a game, with the movements exaggerated on purpose.

Alert / curious

Something has their attention. Nothing has been decided yet.

Conflicted

Two motivations at once, wanting to approach and wanting to withdraw.

Fearful / avoidant

Trying to get smaller, lower, or further away.

Defensive / warning

An explicit request for distance. Communication, not misbehaviour.

Resource guarding

Defending something valued, whether food, an object, a place, or sometimes a person.

Appeasement

Actively signalling non-threat to defuse tension.

Over-aroused

The volume is high, too high for the dog to think clearly.

Possible pain or illness

A pattern with a plausible physical explanation. Worth ruling out first.

The four confidence bands

Confidence is worded, never a fake percentage, and it is capped: the strongest band is still not certainty. Each band is defined by exactly what the code checks, and the reading’s receipt shows the arithmetic.

Insufficient signal

Too little to read. Fewer than two body signals, or no context given. We will not guess from this.

Ambiguous

Two or more readings fit your observation almost equally well. The honest answer is that this could go either way.

Plausible

One reading fits better than the others, but the margin is narrow. Treat it as a working hypothesis and check the next signal.

Consistent

Three or more signals were recorded, at least two point to this reading, it leads by a clear margin, at least one supporting signal has peer-reviewed or scholarly backing, and you gave enough context. The strongest band we issue, and still not certainty.

What counts as a signal, and why context is required

You build an observation from what you can actually see and hear: ears, eyes, mouth and face, tail position, tail movement, body and weight, hackles, displacement behaviours, and an optional sound. None of these is read alone. The engine returns nothing until you have given at least two body signals and the context, where you are and what is happening, because the same posture means different things in a vet’s waiting room and a back garden. Each signal has its own entry in the field guide, and the full method, weights and sources are laid out on the method page.

Privacy

Nothing here leaves your device. There is no account, no upload, and no analytics. Your dog's name and journal live in this browser's local storage, and you can export or delete them at any time from the journal above. A shared link carries only the ticks you made, encoded in the address itself, never your dog's profile or notes.

The site sends Permissions-Policy: microphone=(), camera=(), so your browser blocks microphone access before any script could ask for it. How to verify this yourself.