feat: update Next.js build assets and add cloud setup E2E tests

- Update Next.js generated chunk hashes and build IDs reflecting latest dashboard build
- Update CSS stylesheet references in workflow upload page metadata
- Add comprehensive cloud setup E2E test suite (cloud_setup_test.go) with SSH password/key auth, post-command variable expansion, and Ansible integration
- Fix API priority levels to include 'medium' priority in test coverage
- Add agent-sdk test workflows (minimal, config, codex, multi-agent, session variants)
- Update E2E test utilities with runCLIInBase helper for multi-step cloud config tests
- Fix stderr/stdout capture in dependencies_target_types_test assertions

test/testdata/workflows/agent-and-llm/test-agent-sdk-codex.yaml

@@ -0,0 +1,22 @@
+kind: module
+name: test-agent-sdk-codex
+description: Test agent-sdk step with codex agent
+tags: test,agent-sdk,codex
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  - name: sdk-codex
+    type: agent-sdk
+    log: "Running Codex SDK agent for {{Target}}"
+    agent: codex
+    sdk_config:
+      sandbox: "danger-full-access"
+    messages:
+      - role: user
+        content: "Explain what {{Target}} is."
+    exports:
+      result: "{{sdk_output}}"

test/testdata/workflows/agent-and-llm/test-agent-sdk-config.yaml

@@ -0,0 +1,41 @@
+kind: module
+name: test-agent-sdk-config
+description: Test agent-sdk step with full configuration options
+tags: test,agent-sdk,config
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  # Agent-SDK with cwd, allowed_paths, and sdk_config
+  - name: sdk-configured
+    type: agent-sdk
+    log: "Running configured SDK agent for {{Target}}"
+    agent: claude-code
+    cwd: "{{Output}}"
+    allowed_paths:
+      - "{{Output}}"
+      - "/tmp"
+    sdk_config:
+      model: "sonnet"
+      max_turns: 10
+      env:
+        CUSTOM_VAR: "hello"
+        TARGET_DOMAIN: "{{Target}}"
+    messages:
+      - role: system
+        content: "You are a security reconnaissance agent. Be concise."
+      - role: user
+        content: "Analyze the target {{Target}} and create a summary file."
+    exports:
+      analysis: "{{sdk_output}}"
+      agent_name: "{{sdk_agent}}"
+      session: "{{sdk_session_id}}"
+
+  # Verify exports from previous step
+  - name: verify-sdk-exports
+    type: bash
+    log: "Verifying SDK exports"
+    command: echo "Agent {{agent_name}} produced output (session {{session}})"

test/testdata/workflows/agent-and-llm/test-agent-sdk-minimal.yaml

@@ -0,0 +1,19 @@
+kind: module
+name: test-agent-sdk-minimal
+description: Minimal agent-sdk step test
+tags: test,agent-sdk,quick
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  - name: quick-sdk-agent
+    type: agent-sdk
+    agent: claude-code
+    messages:
+      - role: user
+        content: "Run 'echo hello {{Target}}' and tell me the output."
+    exports:
+      result: "{{sdk_output}}"

test/testdata/workflows/agent-and-llm/test-agent-sdk-multi.yaml

@@ -0,0 +1,54 @@
+kind: module
+name: test-agent-sdk-multi
+description: Test agent-sdk multi-agent strategies (RunFirst / RunAll)
+tags: test,agent-sdk,multi-agent
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  # RunFirst strategy: race multiple agents, return first success
+  - name: sdk-race
+    type: agent-sdk
+    log: "Racing claude-code vs codex for {{Target}}"
+    cwd: "{{Output}}"
+    sdk_config:
+      strategy: "first"
+      agents:
+        - claude-code
+        - codex
+    messages:
+      - role: user
+        content: "Briefly describe what {{Target}} is in one sentence."
+    exports:
+      race_result: "{{sdk_output}}"
+      race_winner: "{{sdk_agent}}"
+
+  - name: show-race-winner
+    type: bash
+    log: "Race winner"
+    command: echo "Winner {{race_winner}} said - {{race_result}}"
+
+  # RunAll strategy: run all agents concurrently, collect all results
+  - name: sdk-all
+    type: agent-sdk
+    log: "Running all agents for {{Target}}"
+    cwd: "{{Output}}"
+    sdk_config:
+      strategy: "all"
+      agents:
+        - claude-code
+        - codex
+    messages:
+      - role: user
+        content: "Give a one-line security assessment of {{Target}}."
+    exports:
+      all_results: "{{sdk_output}}"
+      all_agents: "{{sdk_agent}}"
+
+  - name: show-all-results
+    type: bash
+    log: "All agent results"
+    command: echo "Agents {{all_agents}} produced results"

test/testdata/workflows/agent-and-llm/test-agent-sdk-session.yaml

@@ -0,0 +1,53 @@
+kind: module
+name: test-agent-sdk-session
+description: Test agent-sdk session persistence and resumption (claude-code)
+tags: test,agent-sdk,session
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  # First turn: start a new session
+  - name: sdk-session-start
+    type: agent-sdk
+    log: "Starting new agent session for {{Target}}"
+    agent: claude-code
+    cwd: "{{Output}}"
+    sdk_config:
+      model: "sonnet"
+      max_turns: 5
+    messages:
+      - role: system
+        content: "You are a security analyst. Remember context across turns."
+      - role: user
+        content: "Remember this target: {{Target}}. Reply with OK and the target name."
+    exports:
+      first_output: "{{sdk_output}}"
+      session_id: "{{sdk_session_id}}"
+
+  - name: show-session
+    type: bash
+    log: "Session established"
+    command: echo "Session {{session_id}} - {{first_output}}"
+
+  # Second turn: resume the previous session
+  - name: sdk-session-resume
+    type: agent-sdk
+    log: "Resuming agent session {{session_id}}"
+    agent: claude-code
+    cwd: "{{Output}}"
+    sdk_config:
+      model: "sonnet"
+      session_resume: "{{session_id}}"
+    messages:
+      - role: user
+        content: "What target did I tell you to remember?"
+    exports:
+      resumed_output: "{{sdk_output}}"
+
+  - name: verify-resume
+    type: bash
+    log: "Session resume result"
+    command: echo "Resumed session said - {{resumed_output}}"

test/testdata/workflows/agent-and-llm/test-agent-sdk.yaml

@@ -0,0 +1,27 @@
+kind: module
+name: test-agent-sdk
+description: Test agent-sdk step with go-agent-agnostic library
+tags: test,agent-sdk
+
+params:
+  - name: target
+    required: true
+    default: example.com
+
+steps:
+  # Basic agent-sdk step with built-in agent
+  - name: sdk-basic
+    type: agent-sdk
+    log: "Running SDK agent for {{Target}}"
+    agent: claude-code
+    messages:
+      - role: user
+        content: "List the files in the current directory and summarize the results."
+    exports:
+      sdk_result: "{{sdk_output}}"
+
+  # Use SDK output in a subsequent bash step
+  - name: use-sdk-output
+    type: bash
+    log: "Using SDK agent output"
+    command: echo "Agent said - {{sdk_result}}"