diff --git a/documents/patterns/test-dont-yell.md b/documents/patterns/test-dont-yell.md
new file mode 100644
index 0000000..b7456c3
--- /dev/null
+++ b/documents/patterns/test-dont-yell.md
@@ -0,0 +1,28 @@
+---
+authors: [emmanuel_lin_toulemonde]
+---
+
+# Test, Don't Yell
+
+## Problem
+
+Emphatic instructions ("NEVER do X", "ALWAYS follow Y", "ABSOLUTELY respect this") are unreliable — AI ignores them over long sessions, forgets them as context grows, or complies superficially. Stronger rules don't improve reliability; they just consume context.
+
+## Pattern
+
+Replace emphatic instructions with automated tests or linters AI can run to verify its own output. Instead of telling AI what to do, give it a way to *check* that it did it.
+
+This turns a subjective instruction into a deterministic assertion. AI gets concrete, unambiguous feedback (pass or fail) without relying on memory or interpretation of rules.
+
+**AI can write these checks itself.** Describe the constraint, ask AI to implement the validation. Once implemented, verify it passes on correct code and fails when a violation is introduced. The test suite becomes the living specification.
+
+For every standard or architecture decision, ask: can this be expressed as an automatic validation?
+
+## Example
+
+**Cloud resource name length** — instead of writing `Cloud resource names must be shorter than 63 characters`, add a test that validates all declared resource names. AI runs it and gets a clear failure with the offending name.
+
+Other applications:
+- `Every data product must have a valid data contract` → A test that checks the contract file exists and matches the expected schema.
+- `Always follow this architecture pattern` → An import linter that verifies dependencies only flow in the allowed direction.
+- `Database migrations must be reversible` → A test that runs each migration up then down and checks the schema is unchanged.
diff --git a/documents/relationships.mmd b/documents/relationships.mmd
index d3942d6..345d7fd 100644
--- a/documents/relationships.mmd
+++ b/documents/relationships.mmd
@@ -124,6 +124,12 @@ graph LR
   obstacles/selective-hearing -->|related| obstacles/context-rot
   obstacles/negative-bleedthrough -->|related| obstacles/selective-hearing
   patterns/point-the-target -->|solves| obstacles/negative-bleedthrough
+  patterns/test-dont-yell -->|solves| obstacles/selective-hearing
+  patterns/test-dont-yell -->|solves| anti-patterns/obsess-over-rules
+  patterns/test-dont-yell -->|solves| anti-patterns/perfect-recall-fallacy
+  patterns/test-dont-yell <-->|similar| patterns/habit-hooks
+  patterns/test-dont-yell <-->|similar| patterns/offload-deterministic
+  patterns/test-dont-yell -->|uses| patterns/hooks
 
   %% Obstacle → Anti-pattern relationships (related)
   obstacles/obedient-contractor -->|related| anti-patterns/silent-misalignment
diff --git a/website/config/authors.yaml b/website/config/authors.yaml
index efb3519..8030574 100644
--- a/website/config/authors.yaml
+++ b/website/config/authors.yaml
@@ -23,3 +23,7 @@ rdmueller:
 juan_michelini:
   name: Juan Michelini
   github: juanmichelini
+emmanuel_lin_toulemonde:
+  name: Emmanuel-Lin Toulemonde
+  github: ELToulemonde
+  url: https://eltoulemonde.fr/en