diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index e7e7629825b5f4..685eaf39a2d01a 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -44,6 +44,7 @@ submitted. If you'd like your PR to have the best chance of being merged:
   but features should be confirmed with us first if you aim to avoid wasted
   effort. If there isn't already a GitHub issue for your feature with staff
   confirmation that we want it, start with a GitHub discussion rather than a PR.
+  - This especially applies to any changes proposed to the Zed Extension API.
 - Include a clear description of **what you're solving**, and why it's important.
 - Include **tests**. For UI changes, consider updating visual regression tests (see [Building Zed for macOS](./docs/src/development/macos.md#visual-regression-tests)).
 - If it changes the UI, attach **screenshots** or screen recordings.
@@ -70,6 +71,7 @@ the change with code in hand.
 When your changes affect UI, consult this checklist:
 
 **Accessibility / Ergonomics**
+
 - Do all keyboard shortcuts work as intended?
 - Are shortcuts discoverable (tooltips, menus, docs)?
 - Do all mouse actions work (drag, context menus, resizing, scrolling)?
@@ -78,33 +80,39 @@ When your changes affect UI, consult this checklist:
 - Is it usable without a mouse (keyboard-only navigation)?
 
 **Responsiveness**
+
 - Does the UI scale gracefully on:
-    - Narrow panes (e.g., side-by-side split views)?
-    - Short panes (e.g., laptops with 13" displays)?
-    - High-DPI / Retina displays?
+  - Narrow panes (e.g., side-by-side split views)?
+  - Short panes (e.g., laptops with 13" displays)?
+  - High-DPI / Retina displays?
 - Does resizing panes or windows keep the UI usable and attractive?
 - Do dialogs or modals stay centered and within viewport bounds?
 
 **Platform Consistency**
+
 - Is the feature fully usable on Windows, Linux, and Mac?
 - Does it respect system-level settings (fonts, scaling, input methods)?
 
 **Performance**
+
 - All user interactions must have instant feedback.
-    - If the user requests something slow (e.g. an LLM generation) there should be some indication of the work in progress.
+  - If the user requests something slow (e.g. an LLM generation) there should be some indication of the work in progress.
 - Does it handle large files, big projects, or heavy workloads without degrading?
 - Frames must take no more than 8ms (120fps)
 
 **Consistency**
+
 - Does it match Zed’s design language (spacing, typography, icons)?
 - Are terminology, labels, and tone consistent with the rest of Zed?
 - Are interactions consistent (e.g., how tabs close, how modals dismiss, how errors show)?
 
 **Internationalization & Text**
+
 - Are strings concise, clear, and unambiguous?
 - Do we avoid internal Zed jargon that only insiders would know?
 
 **User Paths & Edge Cases**
+
 - What does the happy path look like?
 - What does the unhappy path look like? (errors, rejections, invalid states)
 - How does it work in offline vs. online states?
@@ -113,17 +121,18 @@ When your changes affect UI, consult this checklist:
 - Are error messages actionable and consistent with Zed’s voice?
 
 **Discoverability & Learning**
+
 - Can a first-time user figure it out without docs?
 - Is there an intuitive way to undo/redo actions?
 - Are power features discoverable but not intrusive?
 - Is there a path from beginner → expert usage (progressive disclosure)?
 
-
 ## Things we will (probably) not merge
 
 Although there are few hard and fast rules, typically we don't merge:
 
 - Anything that can be provided by an extension. For example a new language, or theme. For adding themes or support for a new language to Zed, check out our [docs on developing extensions](https://zed.dev/docs/extensions/developing-extensions).
+- Changes to the Zed Extension API submitted without prior discussion involving Zed staff.
 - New file icons. Zed's default icon theme consists of icons that are hand-designed to fit together in a cohesive manner, please don't submit PRs with off-the-shelf SVGs.
 - Features where (in our subjective opinion) the extra complexity isn't worth it for the number of people who will benefit.
 - Giant refactorings.