Skip to content

Commit 53337bd

Browse files
xingyaowwopenhands-agentryanhoangt
authored
Include 03_browser_use_with_docker_sandboxed_server.py in test-examples workflow (#1021)
Co-authored-by: openhands <[email protected]> Co-authored-by: Hoang Tran <[email protected]>
1 parent e5a7efe commit 53337bd

File tree

6 files changed

+63
-50
lines changed

6 files changed

+63
-50
lines changed

‎.github/workflows/run-examples.yml‎

Lines changed: 33 additions & 36 deletions
Original file line numberDiff line numberDiff line change
@@ -57,9 +57,9 @@ jobs:
5757
# - 08_mcp_with_oauth.py: requires OAuth setup
5858
# - 15_browser_use.py: requires browser setup
5959
# - 16_llm_security_analyzer.py: requires user input
60-
# - 03_browser_use_with_docker_sandboxed_server.py: Docker build cache issues
6160
# - 04_convo_with_api_sandboxed_server.py: requires sandbox API keys
6261
# - 04_vscode_with_docker_sandboxed_server.py: requires VSCode setup
62+
set -e
6363
EXAMPLES=(
6464
"examples/01_standalone_sdk/01_hello_world.py"
6565
"examples/01_standalone_sdk/02_custom_tools.py"
@@ -82,6 +82,7 @@ jobs:
8282
"examples/01_standalone_sdk/24_planning_agent_workflow.py"
8383
"examples/02_remote_agent_server/01_convo_with_local_agent_server.py"
8484
"examples/02_remote_agent_server/02_convo_with_docker_sandboxed_server.py"
85+
"examples/02_remote_agent_server/03_browser_use_with_docker_sandboxed_server.py"
8586
)
8687
8788
# GitHub API setup (only for PR events)
@@ -196,28 +197,38 @@ jobs:
196197
echo ""
197198
echo "Running: $example"
198199
echo "------------------------------------------"
199-
200+
200201
START_TIME=$(date +%s)
201-
202+
202203
# Create temp file to capture output
203204
OUTPUT_FILE=$(mktemp)
204-
205+
205206
# Run example with timeout (20 minutes per example)
206207
# Capture output while still displaying it
207-
if timeout 1200 uv run python "$example" 2>&1 | tee "$OUTPUT_FILE"; then
208-
END_TIME=$(date +%s)
209-
DURATION=$((END_TIME - START_TIME))
210-
DURATION_STR="${DURATION}s"
211-
212-
# Extract cost from output
213-
COST=$(grep "EXAMPLE_COST:" "$OUTPUT_FILE" | awk '{print $2}' | tail -1)
214-
if [ -z "$COST" ]; then
215-
COST="0.00"
216-
fi
217-
218-
# Accumulate total cost
219-
TOTAL_COST=$(echo "$TOTAL_COST + $COST" | bc -l)
220-
208+
# Use || true to prevent script exit on failure
209+
(timeout 1200 uv run python "$example" 2>&1 || true) | tee "$OUTPUT_FILE"
210+
211+
# Check if command succeeded by looking at Python exit
212+
if ! grep -q "EXAMPLE_COST:" "$OUTPUT_FILE"; then
213+
EXIT_CODE=1
214+
else
215+
EXIT_CODE=0
216+
fi
217+
218+
END_TIME=$(date +%s)
219+
DURATION=$((END_TIME - START_TIME))
220+
DURATION_STR="${DURATION}s"
221+
222+
# Extract cost from output
223+
COST=$(grep "EXAMPLE_COST:" "$OUTPUT_FILE" | awk '{print $2}' | tail -1 || echo "0.00")
224+
if [ -z "$COST" ]; then
225+
COST="0.00"
226+
fi
227+
228+
# Accumulate total cost
229+
TOTAL_COST=$(echo "$TOTAL_COST + $COST" | bc -l 2>/dev/null || echo "$TOTAL_COST")
230+
231+
if [ "$EXIT_CODE" -eq 0 ]; then
221232
echo "✓ PASSED: $example (${DURATION_STR}, cost: \$${COST})"
222233
PASSED=$((PASSED + 1))
223234
COMPLETED=$((COMPLETED + 1))
@@ -226,20 +237,6 @@ jobs:
226237
TEST_COST[$example]="$(format_cost $COST)"
227238
echo "PASS|$example|${DURATION}|${COST}" >> "$RESULTS_FILE"
228239
else
229-
EXIT_CODE=$?
230-
END_TIME=$(date +%s)
231-
DURATION=$((END_TIME - START_TIME))
232-
DURATION_STR="${DURATION}s"
233-
234-
# Try to extract cost even for failed tests
235-
COST=$(grep "EXAMPLE_COST:" "$OUTPUT_FILE" | awk '{print $2}' | tail -1)
236-
if [ -z "$COST" ]; then
237-
COST="0.00"
238-
fi
239-
240-
# Accumulate total cost
241-
TOTAL_COST=$(echo "$TOTAL_COST + $COST" | bc -l)
242-
243240
echo "✗ FAILED: $example (exit code: $EXIT_CODE, ${DURATION_STR}, cost: \$${COST})"
244241
FAILED=$((FAILED + 1))
245242
COMPLETED=$((COMPLETED + 1))
@@ -249,13 +246,13 @@ jobs:
249246
TEST_COST[$example]="$(format_cost $COST)"
250247
echo "FAIL|$example|$EXIT_CODE|${DURATION}|${COST}" >> "$RESULTS_FILE"
251248
fi
252-
249+
253250
# Clean up temp file
254251
rm -f "$OUTPUT_FILE"
255-
256-
# Update PR comment after each test
252+
253+
# Update PR comment after each test (with error handling)
257254
echo "Updating PR comment..."
258-
update_comment "$(generate_table)"
255+
update_comment "$(generate_table)" || echo "Warning: Failed to update PR comment"
259256
done
260257
261258
echo ""

‎examples/02_remote_agent_server/01_convo_with_local_agent_server.py‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -238,6 +238,9 @@ def event_callback(event):
238238
logger.info(f" - {event}")
239239

240240
# Report cost (must be before conversation.close())
241+
conversation.state._cached_state = (
242+
None # Invalidate cache to fetch latest stats
243+
)
241244
cost = conversation.conversation_stats.get_combined_metrics().accumulated_cost
242245
print(f"EXAMPLE_COST: {cost}")
243246

‎examples/02_remote_agent_server/02_convo_with_docker_sandboxed_server.py‎

Lines changed: 5 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -40,9 +40,9 @@ def detect_platform():
4040
# the Docker container automatically
4141
with DockerWorkspace(
4242
# dynamically build agent-server image
43-
# base_image="nikolaik/python-nodejs:python3.12-nodejs22",
43+
base_image="nikolaik/python-nodejs:python3.12-nodejs22",
4444
# use pre-built image for faster startup
45-
server_image="ghcr.io/openhands/agent-server:main-python",
45+
# server_image="ghcr.io/openhands/agent-server:main-python",
4646
host_port=8010,
4747
platform=detect_platform(),
4848
) as workspace:
@@ -102,6 +102,9 @@ def event_callback(event) -> None:
102102
logger.info("✅ Second task completed!")
103103

104104
# Report cost (must be before conversation.close())
105+
conversation.state._cached_state = (
106+
None # Invalidate cache to fetch latest stats
107+
)
105108
cost = conversation.conversation_stats.get_combined_metrics().accumulated_cost
106109
print(f"EXAMPLE_COST: {cost}")
107110
finally:

‎examples/02_remote_agent_server/03_browser_use_with_docker_sandboxed_server.py‎

Lines changed: 20 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -34,7 +34,7 @@ def detect_platform():
3434
# Create a Docker-based remote workspace with extra ports for browser access
3535
with DockerWorkspace(
3636
base_image="nikolaik/python-nodejs:python3.12-nodejs22",
37-
host_port=8010,
37+
host_port=8011,
3838
platform=detect_platform(),
3939
extra_ports=True, # Expose extra ports for VSCode and VNC
4040
) as workspace:
@@ -73,14 +73,23 @@ def event_callback(event) -> None:
7373
)
7474
conversation.run()
7575

76-
# Wait for user confirm to exit
77-
y = None
78-
while y != "y":
79-
y = input(
80-
"Because you've enabled extra_ports=True in DockerWorkspace, "
81-
"you can open a browser tab to see the *actual* browser OpenHands "
82-
"is interacting with via VNC.\n\n"
83-
"Link: http://localhost:8012/vnc.html?autoconnect=1&resize=remote\n\n"
84-
"Press 'y' and Enter to exit and terminate the workspace.\n"
85-
">> "
76+
conversation.state._cached_state = None # Invalidate cache to fetch latest stats
77+
cost = conversation.conversation_stats.get_combined_metrics().accumulated_cost
78+
print(f"EXAMPLE_COST: {cost}")
79+
80+
if os.getenv("CI"):
81+
logger.info(
82+
"CI environment detected; skipping interactive prompt and closing workspace." # noqa: E501
8683
)
84+
else:
85+
# Wait for user confirm to exit when running locally
86+
y = None
87+
while y != "y":
88+
y = input(
89+
"Because you've enabled extra_ports=True in DockerWorkspace, "
90+
"you can open a browser tab to see the *actual* browser OpenHands "
91+
"is interacting with via VNC.\n\n"
92+
"Link: http://localhost:8012/vnc.html?autoconnect=1&resize=remote\n\n"
93+
"Press 'y' and Enter to exit and terminate the workspace.\n"
94+
">> "
95+
)

‎openhands-sdk/openhands/sdk/conversation/impl/remote_conversation.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -524,7 +524,7 @@ def state(self) -> RemoteState:
524524
def conversation_stats(self) -> ConversationStats:
525525
"""Get conversation stats from remote server."""
526526
info = self._state._get_conversation_info()
527-
stats_data = info.get("conversation_stats", {})
527+
stats_data = info.get("stats", {})
528528
return ConversationStats.model_validate(stats_data)
529529

530530
@property

‎openhands-workspace/openhands/workspace/docker/workspace.py‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -172,6 +172,7 @@ def model_post_init(self, context: Any) -> None:
172172
base_image=self.base_image,
173173
target=self.target,
174174
platforms=[self.platform],
175+
push=False,
175176
)
176177
tags = build(opts=build_opts)
177178
assert tags and len(tags) > 0, "Build failed, no image tags returned"

0 commit comments

Comments
 (0)