diff --git a/.github/workflows/pr-test.yml b/.github/workflows/pr-test.yml
index c322bdcb..98b238e1 100644
--- a/.github/workflows/pr-test.yml
+++ b/.github/workflows/pr-test.yml
@@ -361,9 +361,9 @@ jobs:
             });
 
   # =============================================================================
-  # AI Tests (agent, aigc) - Run once with SQLite
+  # Agent Tests (agent, aigc) - Run once with SQLite
   # =============================================================================
-  AITest:
+  AgentTest:
     runs-on: ubuntu-latest
     services:
       qdrant:
@@ -448,7 +448,7 @@ jobs:
               owner: context.repo.owner,
               repo: context.repo.repo,
               issue_number: issue_number,
-              body: '🤖 AI Tests (agent, aigc) running with SQLite...'
+              body: '🤖 Agent Tests (agent, aigc) running with SQLite...'
             });
 
       - name: Checkout Kun
@@ -550,20 +550,20 @@ jobs:
           docker pull yaoapp/sandbox-base:latest || true
           docker pull yaoapp/sandbox-claude:latest || true
 
-      - name: Run AI Tests (agent, aigc)
+      - name: Run Agent Tests (agent, aigc)
         env:
           YAO_SANDBOX_WORKSPACE: ${{ runner.temp }}/sandbox/workspace
           YAO_SANDBOX_IPC: ${{ runner.temp }}/sandbox/ipc
         run: |
           export YAO_SANDBOX_CONTAINER_USER="$(id -u):$(id -g)"
-          make unit-test-ai
+          make unit-test-agent
 
       - name: Codecov Report
         uses: codecov/codecov-action@v4
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
 
-      - name: "Comment on PR - AI Tests Done"
+      - name: "Comment on PR - Agent Tests Done"
         uses: actions/github-script@v7
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
@@ -574,13 +574,13 @@ jobs:
               owner: context.repo.owner,
               repo: context.repo.repo,
               issue_number: issue_number,
-              body: '✅ AI Tests (agent, aigc) passed!'
+              body: '✅ Agent Tests (agent, aigc) passed!'
             });
 
   # =============================================================================
-  # Robot E2E Tests (agent/robot/api) - Run TestE2E* with real LLM calls
+  # Robot Tests (all agent/robot/... packages) - Unit + E2E with real LLM calls
   # =============================================================================
-  RobotE2ETest:
+  RobotTest:
     runs-on: ubuntu-latest
     services:
       mcp-everything:
@@ -645,7 +645,7 @@ jobs:
               owner: context.repo.owner,
               repo: context.repo.repo,
               issue_number: issue_number,
-              body: '🤖 Robot E2E Tests running with SQLite...'
+              body: '🤖 Robot Tests (Unit + E2E) running with SQLite...'
             });
 
       - name: Checkout Kun
@@ -725,15 +725,15 @@ jobs:
           echo "YAO_DB_DRIVER=sqlite3" >> $GITHUB_ENV
           echo "YAO_DB_PRIMARY=${{ github.WORKSPACE }}/../app/db/yao.db" >> $GITHUB_ENV
 
-      - name: Run Robot E2E Tests
-        run: make unit-test-robot-e2e
+      - name: Run Robot Tests (Unit + E2E)
+        run: make unit-test-robot
 
       - name: Codecov Report
         uses: codecov/codecov-action@v4
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
 
-      - name: "Comment on PR - Robot E2E Tests Done"
+      - name: "Comment on PR - Robot Tests Done"
         uses: actions/github-script@v7
         with:
           github-token: ${{ secrets.GITHUB_TOKEN }}
@@ -744,7 +744,7 @@ jobs:
               owner: context.repo.owner,
               repo: context.repo.repo,
               issue_number: issue_number,
-              body: '✅ Robot E2E Tests passed!'
+              body: '✅ Robot Tests (Unit + E2E) passed!'
             });
 
   # =============================================================================
diff --git a/.github/workflows/unit-test.yml b/.github/workflows/unit-test.yml
index b81ce354..0442c757 100644
--- a/.github/workflows/unit-test.yml
+++ b/.github/workflows/unit-test.yml
@@ -303,9 +303,9 @@ jobs:
           token: ${{ secrets.CODECOV_TOKEN }}
 
   # =============================================================================
-  # AI Tests (agent, aigc) - Run once with SQLite
+  # Agent Tests (agent, aigc) - Run once with SQLite
   # =============================================================================
-  ai-test:
+  agent-test:
     runs-on: ubuntu-latest
     services:
       qdrant:
@@ -444,13 +444,13 @@ jobs:
           docker pull yaoapp/sandbox-base:latest || true
           docker pull yaoapp/sandbox-claude:latest || true
 
-      - name: Run AI Tests (agent, aigc)
+      - name: Run Agent Tests (agent, aigc)
         env:
           YAO_SANDBOX_WORKSPACE: ${{ runner.temp }}/sandbox/workspace
           YAO_SANDBOX_IPC: ${{ runner.temp }}/sandbox/ipc
         run: |
           export YAO_SANDBOX_CONTAINER_USER="$(id -u):$(id -g)"
-          make unit-test-ai
+          make unit-test-agent
 
       - name: Codecov Report
         uses: codecov/codecov-action@v4
@@ -458,9 +458,9 @@ jobs:
           token: ${{ secrets.CODECOV_TOKEN }}
 
   # =============================================================================
-  # Robot E2E Tests (agent/robot/api) - Run TestE2E* with real LLM calls
+  # Robot Tests (all agent/robot/... packages) - Unit + E2E with real LLM calls
   # =============================================================================
-  robot-e2e-test:
+  robot-test:
     runs-on: ubuntu-latest
     services:
       mcp-everything:
@@ -557,8 +557,8 @@ jobs:
           echo "YAO_DB_DRIVER=sqlite3" >> $GITHUB_ENV
           echo "YAO_DB_PRIMARY=${{ github.WORKSPACE }}/../app/db/yao.db" >> $GITHUB_ENV
 
-      - name: Run Robot E2E Tests
-        run: make unit-test-robot-e2e
+      - name: Run Robot Tests (Unit + E2E)
+        run: make unit-test-robot
 
       - name: Codecov Report
         uses: codecov/codecov-action@v4
diff --git a/Makefile b/Makefile
index cae35a18..a96149b2 100644
--- a/Makefile
+++ b/Makefile
@@ -13,12 +13,12 @@ OS := $(shell uname)
 TESTFOLDER := $(shell $(GO) list ./... | grep -vE 'examples|openai|aigc|neo|twilio|share*' | awk '!/\/tests\// || /openapi\/tests/')
 # Core tests (exclude AI-related: agent, aigc, openai, KB, and sandbox which requires Docker)
 TESTFOLDER_CORE := $(shell $(GO) list ./... | grep -vE 'examples|openai|aigc|neo|twilio|share*|agent|kb|sandbox' | awk '!/\/tests\// || /openapi\/tests/')
-# AI tests (agent, aigc) - exclude agent/search/handlers/web (requires external API keys) and robot/api E2E tests
-TESTFOLDER_AI := $(shell $(GO) list ./agent/... ./aigc/... | grep -v 'agent/search/handlers/web')
+# Agent tests (agent, aigc) - exclude agent/search/handlers/web (requires external API keys) and robot packages (tested in robot job)
+TESTFOLDER_AGENT := $(shell $(GO) list ./agent/... ./aigc/... | grep -vE 'agent/search/handlers/web|agent/robot/')
 # KB tests (kb)
 TESTFOLDER_KB := $(shell $(GO) list ./kb/...)
-# Robot E2E tests (agent/robot/api) - runs TestE2E* tests with real LLM calls
-TESTFOLDER_ROBOT_E2E := $(shell $(GO) list ./agent/robot/api/...)
+# Robot tests (all agent/robot/... packages) - runs ALL tests (unit + E2E) with real LLM calls
+TESTFOLDER_ROBOT := $(shell $(GO) list ./agent/robot/...)
 # Sandbox tests (requires Docker)
 TESTFOLDER_SANDBOX := $(shell $(GO) list ./sandbox/...)
 TESTTAGS ?= ""
@@ -77,12 +77,12 @@ unit-test-core:
 		fi; \
 	done
 
-# AI Unit Test (agent, aigc) - excludes TestE2E* (run separately in unit-test-robot-e2e)
-.PHONY: unit-test-ai
-unit-test-ai:
+# Agent Unit Test (agent, aigc) - excludes robot packages (tested in unit-test-robot) and TestE2E*
+.PHONY: unit-test-agent
+unit-test-agent:
 	echo "mode: count" > coverage.out
-	for d in $(TESTFOLDER_AI); do \
-		$(GO) test -tags $(TESTTAGS) -v -timeout=20m -covermode=count -coverprofile=profile.out -coverpkg=$$(echo $$d | sed "s/\/test$$//g") -skip='TestMemoryLeak|TestIsolateDisposal|TestE2E' $$d > tmp.out; \
+	for d in $(TESTFOLDER_AGENT); do \
+		$(GO) test -tags $(TESTTAGS) -v -timeout=50m -covermode=count -coverprofile=profile.out -coverpkg=$$(echo $$d | sed "s/\/test$$//g") -skip='TestMemoryLeak|TestIsolateDisposal|TestE2E' $$d > tmp.out; \
 		cat tmp.out; \
 		if grep -q "^--- FAIL" tmp.out; then \
 			rm tmp.out; \
@@ -141,13 +141,13 @@ unit-test-kb:
 		fi; \
 	done
 
-# Robot E2E Test (agent/robot/api) - runs TestE2E* tests with real LLM calls
+# Robot Test (all agent/robot/... packages) - runs ALL tests (unit + E2E) with real LLM calls
 # These tests require: LLM API keys, database, and longer timeout
-.PHONY: unit-test-robot-e2e
-unit-test-robot-e2e:
+.PHONY: unit-test-robot
+unit-test-robot:
 	echo "mode: count" > coverage.out
-	for d in $(TESTFOLDER_ROBOT_E2E); do \
-		$(GO) test -tags $(TESTTAGS) -v -timeout=30m -covermode=count -coverprofile=profile.out -coverpkg=$$(echo $$d | sed "s/\/test$$//g") -run='TestE2E' $$d > tmp.out; \
+	for d in $(TESTFOLDER_ROBOT); do \
+		$(GO) test -tags $(TESTTAGS) -v -timeout=50m -covermode=count -coverprofile=profile.out -coverpkg=$$(echo $$d | sed "s/\/test$$//g") -skip='TestMemoryLeak|TestIsolateDisposal' $$d > tmp.out; \
 		cat tmp.out; \
 		if grep -q "^--- FAIL" tmp.out; then \
 			rm tmp.out; \
diff --git a/agent/robot/executor/standard/suspend_test.go b/agent/robot/executor/standard/suspend_test.go
index 4f616ec9..fd746291 100644
--- a/agent/robot/executor/standard/suspend_test.go
+++ b/agent/robot/executor/standard/suspend_test.go
@@ -78,13 +78,12 @@ func TestDetectNeedMoreInfo(t *testing.T) {
 		assert.Empty(t, question)
 	})
 
-	t.Run("handles nested data structure in Next", func(t *testing.T) {
+	t.Run("unwraps data envelope from Next hook", func(t *testing.T) {
 		result := &CallResult{
 			Next: map[string]interface{}{
-				"status":   "need_input",
-				"question": "Which database should I query?",
 				"data": map[string]interface{}{
-					"options": []interface{}{"db1", "db2"},
+					"status":   "need_input",
+					"question": "Which database should I query?",
 				},
 			},
 		}