> ## Documentation Index
> Fetch the complete documentation index at: https://www.agentworldprotocol.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Observations

> Typed, timestamped, multi-channel percepts — pulled per tick or pushed as streams.

Worlds expose named **observation channels**, each declaring a modality, encoding, schema, resolution, and nominal rate in the manifest. An agent subscribes to the subset it can consume.

Every frame has the same shape regardless of what it carries: a per-channel sequence number, a capture timestamp, in streaming a send timestamp, and a payload. The payload is exactly the bytes of the channel's declared modality — a JPEG for an image channel, UTF-8 JSON for a proprioception or event channel — and nothing else.

<Tabs>
  <Tab title="Image">
    | Field        | Value                                                           |
    | ------------ | --------------------------------------------------------------- |
    | `seq`        | 4211                                                            |
    | `ts_mono_ns` | 913004112000                                                    |
    | `payload`    | JPEG bytes (binary), 1280×720 as declared in the channel schema |
  </Tab>

  <Tab title="Proprioception">
    | Field        | Value                                                 |
    | ------------ | ----------------------------------------------------- |
    | `seq`        | 88123                                                 |
    | `ts_mono_ns` | 913004498000                                          |
    | `payload`    | `{"joints_rad": [0.12, -1.04, 0.88], "gripper": 0.5}` |
  </Tab>

  <Tab title="Semantic event">
    | Field        | Value                                                                |
    | ------------ | -------------------------------------------------------------------- |
    | `seq`        | 74                                                                   |
    | `ts_mono_ns` | 913004501000                                                         |
    | `payload`    | `{"event": "entity_appeared", "entity": "box_12", "frame": "world"}` |
  </Tab>
</Tabs>

Delivery depends on the [time model](/concepts/time-models): in **lockstep**, the world sends one frame per channel per tick, usually inline on the control connection; in **streaming**, frames are pushed continuously on their channels with capture and send timestamps and sequence numbers, and the agent measures the staleness of each one against a clock it has synchronized to the world's over the heartbeat. Every frame carries enough metadata (`seq`, `ts_mono_ns`, `ts_send_ns`, channel binding) for the agent to reason about ordering, loss, and latency — see [timestamps and clocks](/spec/semantics/timestamps-and-clocks).