Use default agent preset for integration tests

[simonrosenberg]

Summary

Update integration tests to use the default agent preset from openhands.tools.preset.default so tests validate the same agent configuration shipped to production (GUI/CLI).

Changes:

• BaseIntegrationTest now uses get_default_tools() and get_default_condenser() by default, with an enable_browser property for tests that need browser support
• Functional tests (t01-t08) now inherit default tools from base class, removing redundant tool registration
• t05 and t06 (browsing tests) enable browser via enable_browser property
• t09 keeps custom condenser for token-based condensation testing (special case)
• behavior_helpers.py uses get_default_tools() for behavior tests
• b05 uses default tools from base class

This ensures integration tests validate the exact agent configuration that will be used in production.

Fixes #372

Checklist

• If the PR is changing/adding functionality, are there tests to reflect this?
  • The changes update existing integration tests to use the default preset - no new tests needed as the existing tests now validate the production configuration
• If there is an example, have you run the example to make sure that it works?
  • N/A - this PR modifies test infrastructure
• If there are instructions on how to run the code, have you followed the instructions and made sure that it works?
  • Verified imports work correctly and pre-commit hooks pass
• If the feature is significant enough to require documentation, is there a PR open on the OpenHands/docs repository with the same branch name?
  • N/A - this is an internal test infrastructure change
• Is the github CI passing?
  • Awaiting CI results

@simonrosenberg can click here to continue refining the PR

---

Agent Server images for this PR

• GHCR package: https://github.com/OpenHands/agent-sdk/pkgs/container/agent-server

Variants & Base Images

Variant	Architectures	Base Image	Docs / Tags
java	amd64, arm64	eclipse-temurin:17-jdk	Link
python	amd64, arm64	nikolaik/python-nodejs:python3.12-nodejs22	Link
golang	amd64, arm64	golang:1.21-bookworm	Link

Pull (multi-arch manifest)

# Each variant is a multi-arch manifest supporting both amd64 and arm64
docker pull ghcr.io/openhands/agent-server:d2506a1-python

Run

docker run -it --rm \
  -p 8000:8000 \
  --name agent-server-d2506a1-python \
  ghcr.io/openhands/agent-server:d2506a1-python

All tags pushed for this build

ghcr.io/openhands/agent-server:d2506a1-golang-amd64
ghcr.io/openhands/agent-server:d2506a1-golang_tag_1.21-bookworm-amd64
ghcr.io/openhands/agent-server:d2506a1-golang-arm64
ghcr.io/openhands/agent-server:d2506a1-golang_tag_1.21-bookworm-arm64
ghcr.io/openhands/agent-server:d2506a1-java-amd64
ghcr.io/openhands/agent-server:d2506a1-eclipse-temurin_tag_17-jdk-amd64
ghcr.io/openhands/agent-server:d2506a1-java-arm64
ghcr.io/openhands/agent-server:d2506a1-eclipse-temurin_tag_17-jdk-arm64
ghcr.io/openhands/agent-server:d2506a1-python-amd64
ghcr.io/openhands/agent-server:d2506a1-nikolaik_s_python-nodejs_tag_python3.12-nodejs22-amd64
ghcr.io/openhands/agent-server:d2506a1-python-arm64
ghcr.io/openhands/agent-server:d2506a1-nikolaik_s_python-nodejs_tag_python3.12-nodejs22-arm64
ghcr.io/openhands/agent-server:d2506a1-golang
ghcr.io/openhands/agent-server:d2506a1-java
ghcr.io/openhands/agent-server:d2506a1-python

About Multi-Architecture Support

• Each variant tag (e.g., d2506a1-python) is a multi-arch manifest supporting both amd64 and arm64
• Docker automatically pulls the correct architecture for your platform
• Individual architecture tags (e.g., d2506a1-python-amd64) are also available if needed

[github-actions]

Hi! I started running the integration tests on your PR. You will receive a comment with the results shortly.

[github-actions]

🧪 Integration Tests Results

Overall Success Rate: 96.0%
Total Cost: $2.53
Models Tested: 6
Timestamp: 2026-01-02 18:10:07 UTC

📁 Detailed Logs & Artifacts

Click the links below to access detailed agent/LLM logs showing the complete reasoning process for each model. On the GitHub Actions page, scroll down to the 'Artifacts' section to download the logs.

• litellm_proxy_mistral_devstral_2512: 📥 View & Download Logs
• litellm_proxy_moonshot_kimi_k2_thinking: 📥 View & Download Logs
• litellm_proxy_vertex_ai_gemini_3_pro_preview: 📥 View & Download Logs
• litellm_proxy_deepseek_deepseek_chat: 📥 View & Download Logs
• litellm_proxy_claude_sonnet_4_5_20250929: 📥 View & Download Logs
• litellm_proxy_gpt_5.1_codex_max: 📥 View & Download Logs

📊 Summary

Model	Overall	Integration (Required)	Behavior (Optional)	Tests Passed	Skipped	Total	Cost	Tokens
litellm_proxy_mistral_devstral_2512	75.0%	75.0%	N/A	6/8	1	9	$0.59	1,441,140
litellm_proxy_moonshot_kimi_k2_thinking	100.0%	100.0%	N/A	8/8	1	9	$0.20	304,478
litellm_proxy_vertex_ai_gemini_3_pro_preview	100.0%	100.0%	N/A	9/9	0	9	$0.63	382,177
litellm_proxy_deepseek_deepseek_chat	100.0%	100.0%	N/A	8/8	1	9	$0.06	595,997
litellm_proxy_claude_sonnet_4_5_20250929	100.0%	100.0%	N/A	9/9	0	9	$0.86	865,925
litellm_proxy_gpt_5.1_codex_max	100.0%	100.0%	N/A	8/8	1	9	$0.19	360,650

📋 Detailed Results

litellm_proxy_mistral_devstral_2512

• Overall Success Rate: 75.0% (6/8)
• Integration Tests (Required): 75.0% (6/9)
• Total Cost: $0.59
• Token Usage: prompt: 1,435,188, completion: 5,952
• Run Suffix: litellm_proxy_mistral_devstral_2512_df03e9c_devstral_2512_run_N9_20260102_180259
• Skipped Tests: 1

Skipped Tests:

• t08_image_file_viewing: This test requires a vision-capable LLM model. Please use a model that supports image input.

Failed Tests:

• t02_add_bash_hello ⚠️ REQUIRED: Shell script is not executable (Cost: $0.01)
• t05_simple_browsing ⚠️ REQUIRED: Agent did not find the answer. Response: ... (Cost: $0.37)

litellm_proxy_moonshot_kimi_k2_thinking

• Overall Success Rate: 100.0% (8/8)
• Integration Tests (Required): 100.0% (8/9)
• Total Cost: $0.20
• Token Usage: prompt: 293,534, completion: 10,944, cache_read: 228,352
• Run Suffix: litellm_proxy_moonshot_kimi_k2_thinking_df03e9c_kimi_k2_run_N9_20260102_180259
• Skipped Tests: 1

Skipped Tests:

• t08_image_file_viewing: This test requires a vision-capable LLM model. Please use a model that supports image input.

litellm_proxy_vertex_ai_gemini_3_pro_preview

• Overall Success Rate: 100.0% (9/9)
• Integration Tests (Required): 100.0% (9/9)
• Total Cost: $0.63
• Token Usage: prompt: 359,621, completion: 22,556, cache_read: 198,493, reasoning: 15,864
• Run Suffix: litellm_proxy_vertex_ai_gemini_3_pro_preview_df03e9c_gemini_3_pro_run_N9_20260102_180259

litellm_proxy_deepseek_deepseek_chat

• Overall Success Rate: 100.0% (8/8)
• Integration Tests (Required): 100.0% (8/9)
• Total Cost: $0.06
• Token Usage: prompt: 586,289, completion: 9,708, cache_read: 541,696
• Run Suffix: litellm_proxy_deepseek_deepseek_chat_df03e9c_deepseek_run_N9_20260102_180300
• Skipped Tests: 1

Skipped Tests:

• t08_image_file_viewing: This test requires a vision-capable LLM model. Please use a model that supports image input.

litellm_proxy_claude_sonnet_4_5_20250929

• Overall Success Rate: 100.0% (9/9)
• Integration Tests (Required): 100.0% (9/9)
• Total Cost: $0.86
• Token Usage: prompt: 852,224, completion: 13,701, cache_read: 736,965, cache_write: 114,204, reasoning: 4,193
• Run Suffix: litellm_proxy_claude_sonnet_4_5_20250929_df03e9c_sonnet_run_N9_20260102_180300

litellm_proxy_gpt_5.1_codex_max

• Overall Success Rate: 100.0% (8/8)
• Integration Tests (Required): 100.0% (8/9)
• Total Cost: $0.19
• Token Usage: prompt: 356,730, completion: 3,920, cache_read: 259,840, reasoning: 1,536
• Run Suffix: litellm_proxy_gpt_5.1_codex_max_df03e9c_gpt51_codex_run_N9_20260102_180300
• Skipped Tests: 1

Skipped Tests:

• t09_token_condenser: This test stresses long repetitive tool loops to trigger token-based condensation. GPT-5.1 Codex Max often declines such requests for efficiency/safety reasons.