fix: workflow incorrectly marked as completed while nodes are still executing#138
Open
Conversation
Code Review by Qodo
1. print() in propagate_skip_from_edge
|
![qodo-code-review[bot]](https://avatars.githubusercontent.com/in/484649?s=60&v=4){kind=small}
| if edge_states["has_taken"]: | ||
| # Enqueue node | ||
| # Start execution and enqueue node | ||
| print(f"Starting execution for node: {downstream_node_id}") |
There was a problem hiding this comment.
1. print() in propagate_skip_from_edge 📘 Rule violation ✧ Quality
• The backend module introduces a print() call, which bypasses the standard logging system and reduces production observability. • This violates the requirement to use a module-level logger instead of print(), making output harder to control, route, and structure consistently.
Agent prompt
## Issue description
Non-test backend code uses `print()` instead of the standard logging system.
## Issue Context
Compliance requires a module-level logger (e.g., `logger = logging.getLogger(__name__)`) and forbids `print()` in production backend modules.
## Fix Focus Areas
- api/core/workflow/graph_engine/graph_traversal/skip_propagator.py[5-65]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
+204
to
+219
| def get_incoming_edges_side_effect(node_id): | ||
| if node_id == "node_2": | ||
| return [edge1] | ||
| elif node_id == "node_4": | ||
| return [edge3] | ||
| return [] | ||
|
|
||
| mock_graph.get_incoming_edges.side_effect = get_incoming_edges_side_effect | ||
|
|
||
| # Setup get_outgoing_edges to return different values based on node | ||
| def get_outgoing_edges_side_effect(node_id): | ||
| if node_id == "node_2": | ||
| return [edge3] | ||
| elif node_id == "node_4": | ||
| return [] # No outgoing edges, stops recursion | ||
| return [] |
There was a problem hiding this comment.
2. Untyped side-effect helper functions 📘 Rule violation ✓ Correctness
• The new test file defines local helper functions (get_incoming_edges_side_effect, get_outgoing_edges_side_effect) without parameter and return type annotations. • This violates the requirement that Python function definitions include type hints for all parameters and an explicit return type, reducing type safety and maintainability.
Agent prompt
## Issue description
Local helper functions in the new test file are missing parameter and return type annotations.
## Issue Context
Compliance requires type hints for all function parameters and return values (Python 3.12+ typing style).
## Fix Focus Areas
- api/tests/unit_tests/core/workflow/graph_engine/graph_traversal/test_skip_propagator.py[204-221]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
83
to
+86
| for edge in outgoing_edges: | ||
| self._state_manager.mark_edge_skipped(edge.id) | ||
| # Recursively propagate skip | ||
| self.propagate_skip_from_edge(edge.id) | ||
| self._state_manager.mark_edge_skipped(edge.id) |
There was a problem hiding this comment.
3. Skip propagation order bug 🐞 Bug ✓ Correctness
• _propagate_skip_to_node() now calls propagate_skip_from_edge() before marking the outgoing edge as SKIPPED, but propagate_skip_from_edge() is documented/implemented to operate “from a skipped edge”. • Because edge states default to UNKNOWN and propagate_skip_from_edge() returns early when any incoming edge is UNKNOWN, this reordering can prevent skip propagation from reaching downstream nodes, leaving them stuck in UNKNOWN.
Agent prompt
### Issue description
`_propagate_skip_to_node()` currently calls `propagate_skip_from_edge()` before marking the edge as skipped. But `propagate_skip_from_edge()` assumes it is called *from a skipped edge* and will return early if any incoming edge is `UNKNOWN`. Because edges default to `UNKNOWN`, this can stop skip propagation prematurely.
### Issue Context
The same module (`skip_branch_paths`) marks the edge skipped before propagating, indicating intended behavior.
### Fix Focus Areas
- api/core/workflow/graph_engine/graph_traversal/skip_propagator.py[81-87]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
+63
to
64
| self._state_manager.start_execution(downstream_node_id) | ||
| self._state_manager.enqueue_node(downstream_node_id) |
There was a problem hiding this comment.
4. Start before enqueue 🐞 Bug ⛯ Reliability
• SkipPropagator now calls start_execution() before enqueue_node(), unlike other engine call sites that enqueue first. • If enqueue_node() raises (e.g., missing node id), the executing set will already be mutated, which can keep is_execution_complete() false and leave the system in a partially-updated state.
Agent prompt
### Issue description
`SkipPropagator.propagate_skip_from_edge()` calls `start_execution()` before `enqueue_node()`, unlike other call sites. This can leave partially-updated execution tracking if `enqueue_node()` fails.
### Issue Context
Other engine call sites (event handlers, graph engine startup) enqueue first, then start tracking.
### Fix Focus Areas
- api/core/workflow/graph_engine/graph_traversal/skip_propagator.py[59-65]
- api/core/workflow/graph_engine/graph_state_manager.py[42-55]
- api/core/workflow/graph_engine/event_management/event_handlers.py[194-198]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
+115
to
+150
| def test_propagate_skip_to_node_marks_node_and_outgoing_edges_skipped(self) -> None: | ||
| """_propagate_skip_to_node should mark node and all outgoing edges as skipped.""" | ||
| # Arrange | ||
| mock_graph = create_autospec(Graph) | ||
| mock_state_manager = create_autospec(GraphStateManager) | ||
|
|
||
| # Create outgoing edges | ||
| edge1 = MagicMock(spec=Edge) | ||
| edge1.id = "edge_2" | ||
| edge1.head = "node_downstream_1" # Set head for propagate_skip_from_edge | ||
|
|
||
| edge2 = MagicMock(spec=Edge) | ||
| edge2.id = "edge_3" | ||
| edge2.head = "node_downstream_2" | ||
|
|
||
| # Setup graph edges dict for propagate_skip_from_edge | ||
| mock_graph.edges = {"edge_2": edge1, "edge_3": edge2} | ||
| mock_graph.get_outgoing_edges.return_value = [edge1, edge2] | ||
|
|
||
| # Setup get_incoming_edges to return empty list to stop recursion | ||
| mock_graph.get_incoming_edges.return_value = [] | ||
|
|
||
| propagator = SkipPropagator(mock_graph, mock_state_manager) | ||
|
|
||
| # Use mock to call private method | ||
| # Act | ||
| propagator._propagate_skip_to_node("node_1") | ||
|
|
||
| # Assert | ||
| mock_state_manager.mark_node_skipped.assert_called_once_with("node_1") | ||
| mock_state_manager.mark_edge_skipped.assert_any_call("edge_2") | ||
| mock_state_manager.mark_edge_skipped.assert_any_call("edge_3") | ||
| assert mock_state_manager.mark_edge_skipped.call_count == 2 | ||
| # Should recursively propagate from each edge | ||
| # Since propagate_skip_from_edge is called, we need to verify it was called | ||
| # But we can't directly verify due to recursion. We'll trust the logic. |
There was a problem hiding this comment.
5. Tests don’t verify recursion 🐞 Bug ⛯ Reliability
• Some newly-added tests don’t assert that recursive skip propagation actually occurs; they explicitly state they “trust the logic”. • Several tests rely heavily on mocks and do not set up concrete edge states, reducing the chance they catch regressions in skip propagation behavior.
Agent prompt
### Issue description
Some tests do not verify recursive behavior and rely on mocks without concrete edge state setup, which can mask regressions.
### Issue Context
`propagate_skip_from_edge()` has early-return branches based on `edge_states["has_unknown"]` and other flags; tests should exercise these branches with real edge state values.
### Fix Focus Areas
- api/tests/unit_tests/core/workflow/graph_engine/graph_traversal/test_skip_propagator.py[115-185]
- api/core/workflow/graph_engine/graph_traversal/skip_propagator.py[45-97]
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Benchmark PR from qodo-benchmark#441