fix(agents-api): fix unwrap in prompt step

[Ahmad-mtos]

PR Type

Bug fix, Tests

---

Description

• Fixed logic in _handle_PromptStep to handle unwrap and auto_run_tools conditions.

• Updated test case to reflect changes in _handle_PromptStep.

• Improved code readability by refactoring conditional logic.

---

Changes walkthrough 📝

	Relevant files
Bug fix	__init__.py Refactor `_handle_PromptStep` logic for clarity                    agents-api/agents_api/workflows/task_execution/init.py Refactored conditional logic in _handle_PromptStep. Moved choice assignment after conditional block. Improved readability of unwrap and auto_run_tools checks. +7/-3
Tests	test_task_execution_workflow.py Update test case for `_handle_PromptStep` changes                agents-api/tests/test_task_execution_workflow.py Updated test case to use a string message instead of a dictionary. Adjusted test to align with _handle_PromptStep changes. +1/-1

---

Need help?

Type /help how to ... in the comments thread for any questions about Qodo Merge usage.

Check out the documentation for more information.

---

Important

Fix condition in _handle_PromptStep to correctly handle early return cases and update corresponding test.

• Behavior:
  • Fix condition in _handle_PromptStep in __init__.py to return early with PartialTransition if step.unwrap is true, step.auto_run_tools is false, or finish_reason is not tool_calls.
• Tests:
  • Update test case in test_task_execution_workflow.py to use a string for message instead of a dictionary, aligning with the new logic in _handle_PromptStep.

^{This description was created by for c0e6c66. It will automatically update as commits are pushed.}

[qodo-merge-pro-for-open-source]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 PR contains tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Test Data Mismatch The test case was updated to use a string message "Hello there" but _handle_PromptStep expects a dictionary with "choices" key. This will cause the test to fail when accessing message["choices"]. message = "Hello there" outcome = StepOutcome(output=message) Error Handling No error handling for cases where message dictionary is missing expected keys or has invalid structure. Could raise KeyError when accessing message["choices"] or choice["message"]. message = self.outcome.output if ( step.unwrap or not step.auto_run_tools or message["choices"][0]["finish_reason"] != "tool_calls" ): workflow.logger.debug(f"Prompt step: Received response: {message}") return PartialTransition(output=message) choice = message["choices"][0] tool_calls_input = choice["message"]["tool_calls"]

[qodo-merge-pro-for-open-source]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Score
Possible issue	Match test data with expected format The test message was changed from a dictionary to a string, but the code expects a dictionary with 'choices' key. Update the test message to match the expected format. agents-api/tests/test_task_execution_workflow.py [510-511] -message = "Hello there" +message = {"choices": [{"finish_reason": "stop"}]} outcome = StepOutcome(output=message) Apply this suggestion Suggestion importance[1-10]: 9 Why: Critical bug fix that prevents test failures. The code in _handle_PromptStep expects a dictionary with 'choices' key, but the test provides a string, which would cause runtime errors.	9

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to c0e6c66 in 35 seconds

More details

• Looked at 37 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. agents-api/tests/test_task_execution_workflow.py:510

• Draft comment:
  The message should be a dictionary with a 'choices' key, not a string. This is inconsistent with the expected structure in _handle_PromptStep.
• Reason this comment was not posted:
  Decided after close inspection that this draft comment was likely wrong and/or not actionable:
  The test case is specifically testing the unwrap=True scenario of a prompt step. Looking at the test case, it's verifying that when unwrap is True, the raw message is passed through without any special handling. The test passes with the string message, suggesting that this is actually the intended behavior. The comment seems to be making assumptions about required message structure that don't apply in this unwrap=True case.
  I could be wrong about the intended behavior of unwrap=True. Maybe there are requirements about message structure that are documented elsewhere or enforced in the actual implementation.
  The test clearly shows that with unwrap=True, the message is passed through directly in the assertion on line 518. If there were requirements about message structure, the test would fail. The fact that this test passes suggests this is working as intended.
  The comment should be deleted. The test is specifically for the unwrap=True case where the raw message is passed through, and the string message is valid in this context.

Workflow ID: wflow_2UobB6IFpM0EQY9e

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[qodo-merge-pro-for-open-source]

CI Feedback 🧐

(Feedback updated until commit c0e6c66)

A test triggered by this PR failed. Here is an AI-generated analysis of the failure:

Action: Typecheck

Failed stage: Typecheck [❌]

Failed test name: agents_api.common.protocol.tasks

Failure summary: The pytype static type checker failed due to two errors in the file agents_api/common/protocol/tasks.py: Line 209: Type error where list_execution_transitions object is being called but is not callable Line 210: Attribute error trying to access .id on a None value in self.execution_input.execution.id

Relevant error logs: 1: ##[group]Operating System 2: Ubuntu ... 1222: [13/299] check agents_api.autogen.Entries 1223: [14/299] check agents_api.autogen.Tasks 1224: [15/299] check agents_api.autogen.Agents 1225: [16/299] check agents_api.common.utils.datetime 1226: [17/299] check agents_api.common.protocol.developers 1227: [18/299] check agents_api.common.utils.db_exceptions 1228: [19/299] check agents_api.common.protocol.agents 1229: [20/299] check agents_api.queries.utils 1230: ERROR:pytype.matcher Invalid type: <class 'pytype.abstract.function.ParamSpecMatch'> ... 1332: [122/299] check agents_api.queries.docs.__init__ 1333: [123/299] check agents_api.activities.utils 1334: [124/299] check agents_api.queries.developers.__init__ 1335: [125/299] check agents_api.queries.agents.__init__ 1336: [126/299] check agents_api.common.protocol.remote 1337: [127/299] check agents_api.queries.__init__ 1338: [128/299] check agents_api.clients.integrations 1339: [129/299] check agents_api.common.protocol.tasks 1340: FAILED: /home/runner/work/julep/julep/agents-api/.pytype/pyi/agents_api/common/protocol/tasks.pyi 1341: /home/runner/work/julep/julep/agents-api/.venv/bin/python -m pytype.main --disable pyi-error --imports_info /home/runner/work/julep/julep/agents-api/.pytype/imports/agents_api.common.protocol.tasks.imports --module-name agents_api.common.protocol.tasks --platform linux -V 3.12 -o /home/runner/work/julep/julep/agents-api/.pytype/pyi/agents_api/common/protocol/tasks.pyi --analyze-annotated --nofail --none-is-not-bool --quick --strict-none-binding /home/runner/work/julep/julep/agents-api/agents_api/common/protocol/tasks.py 1342: /home/runner/work/julep/julep/agents-api/agents_api/common/protocol/tasks.py:209:29: error: in get_inputs: 'agents_api.queries.executions.list_execution_transitions' object is not callable [not-callable] ... 1345: execution_id=self.execution_input.execution.id, 1346: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1347: limit=1000, 1348: ~~~~~~~~~~~~~~~~~~~~~~~ 1349: direction="asc", 1350: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1351: ) 1352: ~~~~~~~~~ 1353: /home/runner/work/julep/julep/agents-api/agents_api/common/protocol/tasks.py:210:26: error: in get_inputs: No attribute 'id' on None [attribute-error] 1354: In Optional[agents_api.autogen.Executions.Execution] 1355: execution_id=self.execution_input.execution.id, 1356: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1357: For more details, see https://google.github.io/pytype/errors.html ... 1457: [229/299] check agents_api.queries.chat.__init__ 1458: [230/299] check tests.sample_tasks.test_find_selector 1459: [231/299] check tests.__init__ 1460: [232/299] check tests.test_nlp_utilities 1461: [233/299] check agents_api.metrics.__init__ 1462: [234/299] check agents_api.common.exceptions.sessions 1463: [235/299] check agents_api.clients.__init__ 1464: [236/299] check agents_api.rec_sum.entities 1465: ninja: build stopped: cannot make progress due to previous errors. 1466: Computing dependencies 1467: Generated API key since not set in the environment: 72670854398539784241069788237269 1468: Sentry DSN not found. Sentry will not be enabled. 1469: Analyzing 289 sources with 0 local dependencies 1470: Leaving directory '.pytype' 1471: ##[error]Process completed with exit code 1.