diff --git a/.gemini/skills/async-pr-review/SKILL.md b/.gemini/skills/async-pr-review/SKILL.md
index 656a19260f..6db732879e 100644
--- a/.gemini/skills/async-pr-review/SKILL.md
+++ b/.gemini/skills/async-pr-review/SKILL.md
@@ -7,6 +7,10 @@ description: Trigger this skill when the user wants to start an asynchronous PR
 
 This skill provides a set of tools to asynchronously review a Pull Request. It will create a background job to run the project's preflight checks, execute Gemini-powered test plans, and perform a comprehensive code review using custom prompts.
 
+This is a prime example of composing Gemini CLI features:
+1.  **Headless Automation (`async-review.sh`)**: The script invokes `gemini` with `--approval-mode=yolo` and the `/review-frontend` custom command in the background.
+2.  **Agentic Evaluation (`check-async-review.sh`)**: The check script outputs clean JSON/text statuses for the agent to parse. The agent itself synthesizes the final assessment dynamically from the generated log files.
+
 ## Workflow
 
 1.  **Determine Action**: Establish whether the user wants to start a new async review or check the status of an existing one.
@@ -33,4 +37,7 @@ If the user wants to check the status or view the final assessment of a previous
     ```bash
     .gemini/skills/async-pr-review/scripts/check-async-review.sh <PR_NUMBER>
     ```
-3.  This script will launch a live status dashboard. Once all tasks are complete, it will automatically query a new instance of Gemini for a final assessment and provide a recommendation on whether to approve the PR. Let the script run in the user's terminal.
\ No newline at end of file
+3.  **Evaluate Output**: Read the output from the script.
+    *   If the output contains `STATUS: IN_PROGRESS`, tell the user which tasks are still running.
+    *   If the output contains `STATUS: COMPLETE`, use your file reading tools (`read_file`) to retrieve the contents of the `review.md` and `test-execution.log` files from the `LOG_DIR` specified in the output.
+    *   **Final Assessment**: Read those files, synthesize their results, and give the user a concise recommendation on whether the PR builds successfully, passes tests, and if you recommend they approve it based on the review.
\ No newline at end of file
diff --git a/.gemini/skills/async-pr-review/scripts/check-async-review.sh b/.gemini/skills/async-pr-review/scripts/check-async-review.sh
index b9ea7eede8..36ca12aea4 100755
--- a/.gemini/skills/async-pr-review/scripts/check-async-review.sh
+++ b/.gemini/skills/async-pr-review/scripts/check-async-review.sh
@@ -7,87 +7,49 @@ if [[ -z "$pr_number" ]]; then
 fi
 
 log_dir="$HOME/dev/pr-$pr_number/logs"
-GEMINI_CMD="$HOME/.gcli/nightly/node_modules/.bin/gemini"
 
 if [[ ! -d "$log_dir" ]]; then
+  echo "STATUS: NOT_FOUND"
   echo "❌ No logs found for PR #$pr_number in $log_dir"
-  exit 1
+  exit 0
 fi
 
-# Define the tasks: name|log_file
 tasks=(
   "preflight|preflight.log"
   "review|review.md"
   "test-execution|test-execution.log"
 )
 
-while true; do
-  clear
-  echo "📊 Status for PR #$pr_number in $log_dir"
-  echo "================================================="
+all_done=true
+echo "STATUS: CHECKING"
 
-  all_done=true
-
-  for task_info in "${tasks[@]}"; do
-    IFS="|" read -r task_name log_file <<< "$task_info"
-    
-    file_path="$log_dir/$log_file"
-    exit_file="$log_dir/$task_name.exit"
-
-    if [[ -f "$exit_file" ]]; then
-      # Task is done
-      exit_code=$(cat "$exit_file")
-      if [[ "$exit_code" == "0" ]]; then
-        status="✅ SUCCESS"
-      else
-        status="❌ FAILED (exit code $exit_code)"
-      fi
-    elif [[ -f "$file_path" ]]; then
-      status="⏳ RUNNING"
-      all_done=false
-    else
-      status="➖ NOT STARTED"
-      all_done=false
-    fi
-
-    echo "$status - $task_name"
-    
-    if [[ -f "$file_path" ]]; then
-      if [[ "$status" == "⏳ RUNNING" ]]; then
-        # Show what it's currently doing
-        echo "  Last output:"
-        tail -n 3 "$file_path" | sed 's/^/    | /'
-      elif [[ "$status" == *"FAILED"* ]]; then
-        # Show the last lines of the error to help debug
-        echo "  Error snippet:"
-        tail -n 5 "$file_path" | sed 's/^/    | /'
-      fi
-    fi
-    echo ""
-  done
-
-  if $all_done; then
-    break
-  fi
+for task_info in "${tasks[@]}"; do
+  IFS="|" read -r task_name log_file <<< "$task_info"
   
-  echo "⏳ Waiting 5 seconds before checking again..."
-  sleep 5
+  file_path="$log_dir/$log_file"
+  exit_file="$log_dir/$task_name.exit"
+
+  if [[ -f "$exit_file" ]]; then
+    exit_code=$(cat "$exit_file")
+    if [[ "$exit_code" == "0" ]]; then
+      echo "✅ $task_name: SUCCESS"
+    else
+      echo "❌ $task_name: FAILED (exit code $exit_code)"
+      echo "   Last lines of $file_path:"
+      tail -n 3 "$file_path" | sed 's/^/      /'
+    fi
+  elif [[ -f "$file_path" ]]; then
+    echo "⏳ $task_name: RUNNING"
+    all_done=false
+  else
+    echo "➖ $task_name: NOT STARTED"
+    all_done=false
+  fi
 done
 
-echo "🎉 All tasks are complete!"
-echo "🤖 Asking Gemini for final PR assessment..."
-echo "================================================="
-
-"$GEMINI_CMD" "I have just completed async tasks for PR $pr_number. 
-
-Here is the code review output:
-\`\`\`markdown
-$(cat "$log_dir/review.md" 2>/dev/null)
-\`\`\`
-
-Here is the test execution log:
-\`\`\`
-$(cat "$log_dir/test-execution.log" 2>/dev/null)
-\`\`\`
-
-Please evaluate the results. Tell me if the PR builds successfully, if it passes tests, and if you recommend I approve it based on the review. Keep your answer concise and actionable."
+if $all_done; then
+  echo "STATUS: COMPLETE"
+  echo "LOG_DIR: $log_dir"
+else
+  echo "STATUS: IN_PROGRESS"
+fi
\ No newline at end of file