diff --git a/README.md b/README.md
index 0e72b2670..1f8da4916 100644
--- a/README.md
+++ b/README.md
@@ -35,7 +35,7 @@ Begin the "start here" guide for a new user.
 [natural-language inference page](https://pyautofit.readthedocs.io/en/latest/overview/natural_language.html)
 walks through this, including its **AI First Design**.
 
-## Bring Your Own Likelihood
+## Bring Your Own Likelihood (BYOL)
 
 Already have a likelihood function for your science problem? **Point the assistant at your existing code and it can
 set it up with PyAutoFit** — defining the model, choosing priors with you, configuring a search and organising the
@@ -56,6 +56,12 @@ Once inference is running, explain how the results are written to disk and show
 them with PyAutoFit.
 ```
 
+The assistant answers this prompt with its
+[**BYOL** mode](https://github.com/PyAutoLabs/autofit_assistant/blob/main/modes/byol.md) — bring your own likelihood,
+the assistant brings the inference. It reads your code and restates what your likelihood scores, composes a model with
+priors chosen with you, wraps and validates your function unchanged, recommends a search, and runs nothing until you
+say go.
+
 Your existing science code remains the source of the likelihood. With PyAutoFit built around it, you can perform
 inference through natural language while gaining access to features such as flexible priors and model composition,
 MCMC and nested sampling, automated result handling, model comparison and scalable workflows.
diff --git a/docs/overview/natural_language.md b/docs/overview/natural_language.md
index e04a4c011..774669524 100644
--- a/docs/overview/natural_language.md
+++ b/docs/overview/natural_language.md
@@ -92,6 +92,9 @@ For your own project, you can instead ask:
 > the same parameter values.
 :::
 
+This is the job of the assistant's [**BYOL** mode](https://github.com/PyAutoLabs/autofit_assistant/blob/main/modes/byol.md),
+which wraps your existing likelihood unchanged and validates it.
+
 :::{container} ai-first-design
 **AI First Design:** PyAutoFit gives the agent a small, testable integration task: connect named model parameters to your existing likelihood and check that its numerical outputs are unchanged. Your validated science code then becomes available to PyAutoFit's searches and result-analysis tools, without the agent having to reimplement it.
 :::
diff --git a/docs/overview/quick_start.md b/docs/overview/quick_start.md
index 616413f4c..332fef21b 100644
--- a/docs/overview/quick_start.md
+++ b/docs/overview/quick_start.md
@@ -38,7 +38,7 @@ six steps on your own science. At any step you can ask a question (for example
 "what is a prior?") or say "teacher mode" to get a full explanation of everything
 as you go. The six steps are the sections of [Natural Language Inference](https://pyautofit.readthedocs.io/en/latest/overview/natural_language.html).
 
-## Bring Your Own Likelihood
+## Bring Your Own Likelihood (BYOL)
 
 Already have a likelihood function for your science problem? **Point the
 assistant at your existing code and it can set it up with PyAutoFit** —
@@ -60,6 +60,13 @@ organising the results:
 > show me how to inspect and interpret them with PyAutoFit.
 :::
 
+The assistant answers this prompt with its
+[**BYOL** mode](https://github.com/PyAutoLabs/autofit_assistant/blob/main/modes/byol.md) — bring your
+own likelihood, the assistant brings the inference. It reads your code and
+restates what your likelihood scores, composes a model with priors chosen
+with you, wraps and validates your function unchanged, recommends a search,
+and runs nothing until you say go.
+
 Your existing science code remains the source of the likelihood. With
 PyAutoFit built around it, you can perform inference through natural
 language while gaining access to features such as flexible priors and model