Update 0278-env-context.md (#299)

docs/adrs/0278-env-context.md

@@ -15,6 +15,7 @@ User wants to reference workflow variables defined in workflow yaml file for act
 Runner will create and populate the `env` context for every job execution using following logic:
 1. On job start, create `env` context with any environment variables in the job message, these are env defined in customer's YAML file's job/workflow level `env` section.
 2. Update `env` context when customer use `::set-env::` to set env at the runner level.
+3. Update `env` context with step's `env` block before each step runs.
 
 The `env` context is only available in the runner, customer can't use the `env` context in any server evaluation part, just like the `runner` context
 
@@ -22,17 +23,25 @@ Example yaml:
 ```yaml
 
 env:
-  env1: 100
+  env1: 10
+  env2: 20
+  env3: 30
 jobs:
   build:
     env:
+      env1: 100
       env2: 200
     runs-on: ubuntu-latest
     steps:
       - run: |
-          echo ${{ env.env1 }}
+          echo ${{ env.env1 }}  // 1000
-        if: env.env2 == 200
+          echo $env1            // 1000
-        name: ${{ env.env1 }}_${{ env.env2 }}
+          echo $env2            // 200
+          echo $env3            // 30
+        if: env.env2 == 200     // true
+        name: ${{ env.env1 }}_${{ env.env2 }}   //1000_200
+        env:
+          env1: 1000
 ```
 
 ### Don't populate the `env` context with environment variables from runner machine. 
@@ -48,4 +57,4 @@ build:
   - uses: docker://ubuntu:18.04
     with:
       args: echo ${{env.USER}}  <- what should customer expect this output? runner/root
-```
+```