Skip to content

The cross-harness, self-improving memory layer for AI agents. Join our community: https://discord.com/invite/zdNChS2fBu

docs: add architecture summary and docs links to README

Claude2d8ebcf3d507
diff --git a/README.md b/README.md
index 682f5e7c..1983cc0b 100644
--- a/README.md
+++ b/README.md
@@ -51,6 +51,10 @@ Learn more in the [Agent Beacon documentation](https://docs.asymptotelabs.ai).
## High-Level Architecture
+Beacon captures activity where each agent actually runs, then normalizes it on the
+endpoint into a single OpenTelemetry-based event model. That one stream is what the
+local dashboard, JSONL retention, and every SIEM destination read from.
+
<p align="center">
<img src="images/beacon-architecture.png" alt="Beacon endpoint architecture" width="860">
</p>
@@ -63,7 +67,11 @@ Learn more in the [Agent Beacon documentation](https://docs.asymptotelabs.ai).
records into the [major enterprise-grade SIEMs](#output-destinations).
Collection, processing, and inspection stay local by default; the same normalized
-event model extends to CI and cloud-agent paths under customer control.
+event model extends to CI and cloud-agent paths under customer control. See the
+[open-source architecture reference](https://docs.asymptotelabs.ai/architecture/architecture)
+for the full collection, normalization, storage, and forwarding pipeline, or the
+[system architecture overview](https://docs.asymptotelabs.ai/architecture/system-architecture)
+to compare it with the managed path.
## Supported Surfaces