diff --git a/docs/components/indexers/editorial-content.md b/docs/components/indexers/editorial-content.md
index e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..e9f31c3987b01e6c61cfd89599a00f74201a7e4b 100644
--- a/docs/components/indexers/editorial-content.md
+++ b/docs/components/indexers/editorial-content.md
@@ -0,0 +1,30 @@
+# Editorial Content Indexer
+
+## Features
+
+This indexer provides methods to index in Elasticsearch the editorial content (pages, posts) that is produced with the Ghost CMS. It relies on Ghost's "Custom Integrations" feature, cf. https://ghost.org/docs/api/v2/webhooks/.
+
+It provides two indexation strategies:
+* full
+* incremental (adding/updating/deleting pages or posts)
+
+
+As an additional feature, it includes in the Elasticsearch index the *relations* of a given post or page to other posts, pages and datasets.
+
+
+## Requirements
+
+* A working Ghost installation
+* An Admin API Key (provided by Ghost, when configuring a custom integration)
+* A working Elasticsearch installation
+
+
+## Implementation
+
+The indexer is developed in Python, using the Flask framework to expose an API. The authentified access to the Ghost API requires a JWT which is generated with the [JSON Web Token JavaScript module](https://www.npmjs.com/package/jsonwebtoken)
+
+## TODO
+* see whether the JWT could be generated in Python
+* understand why a `get_content` is performed despite the content being pushed when the event is fired
+* implement the support for multiple languages (EN, FR, ...)
+* exploit logs
\ No newline at end of file
diff --git a/docs/components/off-the-shelf-apps/api-gateway.md b/docs/components/off-the-shelf-apps/api-gateway.md
index cb6977bf32376ba469a5dfcf308acd4b818c96be..2fe1e2ff0cad7f161b9540c09658811d4b56aefd 100644
--- a/docs/components/off-the-shelf-apps/api-gateway.md
+++ b/docs/components/off-the-shelf-apps/api-gateway.md
@@ -1,6 +1,28 @@
 # API Gateway
 
-### TODO
+## Why Kong ?
 
-* Explain the reasons why Kong was chosen
-* Explain which features are used, in particular the AUTHN layer
+* it is Open Source
+* based on Nginx, a widely known Web Server / Reverse Proxy
+* it can be administrated via a dedicated admin REST API
+* it can be deployed in a High Availability fashion
+* it can be extended by various plugins; a Plugin Development Kit is provided
+* it provides several interesting features: load balancing (with active and passive healthcheck), 
+
+## Features
+
+* Load balancer with active healthchecks
+* Plugins : 
+    * jwt
+    * cors
+    * bot-detection (never tested!)
+    * prometheus
+    * key-auth, in order to let (micro-)services authenticate themselves with respect to other (micro-)services
+    * request-transformer, in order to issue authenticated requests towards Matomo
+    * acl, in order to:
+        - let (micro-)service use some potentially dangerous features exposed by other (micro-)services
+        - let the back-office team preview draft posts and access the Admin GUI
+
+## Setup
+
+A Python script allows the administrator to set up the proper configuration, according to the various environments.
diff --git a/docs/miscellaneous/accessibility.md b/docs/miscellaneous/accessibility.md
index 31c413a0ec1fd9a2972fa98c558cc83ba2d315de..e5238072ebe413b96f5caf23e0147040623dbe13 100644
--- a/docs/miscellaneous/accessibility.md
+++ b/docs/miscellaneous/accessibility.md
@@ -21,9 +21,8 @@ pa11y https://data.beta.grandlyon.com/fr/accueil --ignore "warning;notice"
 npm install -g pa11y-ci
 ```
 
-Then create a **.pa11y1** json file.
-
-```bash
+Then create a **.pa11yci** json file.
+```
 {
   "urls": [
     "https://data.beta.grandlyon.com/fr/accueil"