docs: update sandbox persistence information

.github/workflows/test.yml

@@ -19,9 +19,13 @@ jobs:
           require: write
           username: ${{ github.triggering_actor }}
           error-if-missing: true
+        # Skip this check for bot users
+        if: ${{ !endsWith(github.triggering_actor, '[bot]') }}
 
   unit-tests:
+    # Run if access-check is successful or skipped (for bot users)
     needs: access-check
+    if: ${{ always() && (needs.access-check.result == 'success' || needs.access-check.result == 'skipped') }}
     runs-on: ubuntu-latest-8
     steps:
       - uses: actions/checkout@v4
@@ -50,15 +54,15 @@ jobs:
   codemod-tests:
     needs: access-check
     # TODO: re-enable when this check is a develop required check
-    if: false
+    if: ${{ always() && (needs.access-check.result == 'success' || needs.access-check.result == 'skipped') && false }}
     runs-on: ubuntu-latest-32
     strategy:
       matrix:
         sync_graph: [ true, false ]
         size: [ small, large ]
         exclude:
           # Exclude large codemod tests when not needed
-          - size: ${{(contains(github.event.pull_request.labels.*.name, 'big-codemod-tests') || github.event_name == 'push' || github.event_name == 'workflow_dispatch') && 'kevin' || 'large'}}
+          - size: ${{(contains(github.event.pull_request.labels.*.name, 'big-codemod-tests') || github.event_name == 'push' || github.event_name == 'workflow_dispatch') && 'kevin' || 'large'}
           - size: large
             sync_graph: true
     concurrency:
@@ -91,7 +95,7 @@ jobs:
 
   parse-tests:
     needs: access-check
-    if: contains(github.event.pull_request.labels.*.name, 'parse-tests') || github.event_name == 'push' || github.event_name == 'workflow_dispatch'
+    if: ${{ always() && (needs.access-check.result == 'success' || needs.access-check.result == 'skipped') && (contains(github.event.pull_request.labels.*.name, 'parse-tests') || github.event_name == 'push' || github.event_name == 'workflow_dispatch') }}
     runs-on: ubuntu-latest-32
     steps:
       - uses: actions/checkout@v4
@@ -162,6 +166,7 @@ jobs:
 
   integration-tests:
     needs: access-check
+    if: ${{ always() && (needs.access-check.result == 'success' || needs.access-check.result == 'skipped') }}
     runs-on: ubuntu-latest-16
     steps:
       - uses: actions/checkout@v4

docs/sandboxes/overview.mdx

@@ -29,6 +29,5 @@ Agents utilize sandboxes for a variety of tasks:
 Sandboxes are typically configured per-agent run or defined within your Codegen settings. You often don't need to manage them directly, but advanced configurations might allow specifying Docker images or environment variables.
 
 <Note>
-  Sandboxes are ephemeral; their state is not typically persisted between agent
-  runs unless specifically configured for caching or stateful operations.
+  Sandboxes maintain file system persistence between agent interactions within the same context. For example, when continuing a conversation across different Slack messages or Linear comments, the sandbox state is preserved, allowing agents to seamlessly continue their work without losing context or having to reinstall dependencies.
 </Note>{" "}