docs(mintlify): remove OpenAPI Endpoints groups; add localized MDX endpoint pages for pt-BR and ko

feat: add RAG system foundation with generic vector store support

- Add BaseClient protocol for vector stores
- Move BaseRAGStorage to rag/core
- Centralize embedding types in embeddings/types.py
- Remove unused storage models

docs/docs.json

@@ -341,11 +341,12 @@
             "groups": [
               {
                 "group": "Getting Started",
-                "pages": ["en/api-reference/introduction"]
-              },
-              {
-                "group": "Endpoints",
-                "openapi": "https://raw.githubusercontent.com/crewAIInc/crewAI/main/docs/enterprise-api.en.yaml"
+                "pages": [
+                  "en/api-reference/introduction",
+                  "en/api-reference/inputs",
+                  "en/api-reference/kickoff",
+                  "en/api-reference/status"
+                ]
               }
             ]
           },
@@ -680,11 +681,12 @@
             "groups": [
               {
                 "group": "Começando",
-                "pages": ["pt-BR/api-reference/introduction"]
-              },
-              {
-                "group": "Endpoints",
-                "openapi": "https://raw.githubusercontent.com/crewAIInc/crewAI/main/docs/enterprise-api.pt-BR.yaml"
+                "pages": [
+                  "pt-BR/api-reference/introduction",
+                  "pt-BR/api-reference/inputs",
+                  "pt-BR/api-reference/kickoff",
+                  "pt-BR/api-reference/status"
+                ]
               }
             ]
           },
@@ -1026,11 +1028,12 @@
             "groups": [
               {
                 "group": "시작 안내",
-                "pages": ["ko/api-reference/introduction"]
-              },
-              {
-                "group": "Endpoints",
-                "openapi": "https://raw.githubusercontent.com/crewAIInc/crewAI/main/docs/enterprise-api.ko.yaml"
+                "pages": [
+                  "ko/api-reference/introduction",
+                  "ko/api-reference/inputs",
+                  "ko/api-reference/kickoff",
+                  "ko/api-reference/status"
+                ]
               }
             ]
           },
@@ -1081,6 +1084,10 @@
     "indexing": "all"
   },
   "redirects": [
+    {
+      "source": "/api-reference",
+      "destination": "/en/api-reference/introduction"
+    },
     {
       "source": "/introduction",
       "destination": "/en/introduction"
@@ -1133,6 +1140,18 @@
       "source": "/api-reference/:path*",
       "destination": "/en/api-reference/:path*"
     },
+    {
+      "source": "/en/api-reference",
+      "destination": "/en/api-reference/introduction"
+    },
+    {
+      "source": "/pt-BR/api-reference",
+      "destination": "/pt-BR/api-reference/introduction"
+    },
+    {
+      "source": "/ko/api-reference",
+      "destination": "/ko/api-reference/introduction"
+    },
     {
       "source": "/examples/:path*",
       "destination": "/en/examples/:path*"

docs/en/api-reference/inputs.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /inputs"
+description: "Get required inputs for your crew"
+openapi: "/enterprise-api.en.yaml GET /inputs"
+---
+
+

docs/en/api-reference/kickoff.mdx

@@ -0,0 +1,7 @@
+---
+title: "POST /kickoff"
+description: "Start a crew execution"
+openapi: "/enterprise-api.en.yaml POST /kickoff"
+---
+
+

docs/en/api-reference/status.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /status/{kickoff_id}"
+description: "Get execution status"
+openapi: "/enterprise-api.en.yaml GET /status/{kickoff_id}"
+---
+
+

docs/en/concepts/training.mdx

@@ -21,13 +21,17 @@ To use the training feature, follow these steps:
 3. Run the following command:
 
 ```shell
-crewai train -n <n_iterations> <filename> (optional)
+crewai train -n <n_iterations> -f <filename.pkl>
 ```
 <Tip>
   Replace `<n_iterations>` with the desired number of training iterations and `<filename>` with the appropriate filename ending with `.pkl`.
 </Tip>
 
-### Training Your Crew Programmatically
+<Note>
+  If you omit `-f`, the output defaults to `trained_agents_data.pkl` in the current working directory. You can pass an absolute path to control where the file is written.
+</Note>
+
+### Training your Crew programmatically
 
 To train your crew programmatically, use the following steps:
 
@@ -51,19 +55,65 @@ except Exception as e:
     raise Exception(f"An error occurred while training the crew: {e}")
 ```
 
-### Key Points to Note
+## How trained data is used by agents
 
-- **Positive Integer Requirement:** Ensure that the number of iterations (`n_iterations`) is a positive integer. The code will raise a `ValueError` if this condition is not met.
-- **Filename Requirement:** Ensure that the filename ends with `.pkl`. The code will raise a `ValueError` if this condition is not met.
-- **Error Handling:** The code handles subprocess errors and unexpected exceptions, providing error messages to the user.
+CrewAI uses the training artifacts in two ways: during training to incorporate your human feedback, and after training to guide agents with consolidated suggestions.
 
-It is important to note that the training process may take some time, depending on the complexity of your agents and will also require your feedback on each iteration.
+### Training data flow
 
-Once the training is complete, your agents will be equipped with enhanced capabilities and knowledge, ready to tackle complex tasks and provide more consistent and valuable insights.
+```mermaid
+flowchart TD
+  A["Start training<br/>CLI: crewai train -n -f<br/>or Python: crew.train(...)"] --> B["Setup training mode<br/>- task.human_input = true<br/>- disable delegation<br/>- init training_data.pkl + trained file"]
 
-Remember to regularly update and retrain your agents to ensure they stay up-to-date with the latest information and advancements in the field.
+  subgraph "Iterations"
+    direction LR
+    C["Iteration i<br/>initial_output"] --> D["User human_feedback"]
+    D --> E["improved_output"]
+    E --> F["Append to training_data.pkl<br/>by agent_id and iteration"]
+  end
 
-Happy training with CrewAI! 🚀
+  B --> C
+  F --> G{"More iterations?"}
+  G -- "Yes" --> C
+  G -- "No" --> H["Evaluate per agent<br/>aggregate iterations"]
+
+  H --> I["Consolidate<br/>suggestions[] + quality + final_summary"]
+  I --> J["Save by agent role to trained file<br/>(default: trained_agents_data.pkl)"]
+
+  J --> K["Normal (non-training) runs"]
+  K --> L["Auto-load suggestions<br/>from trained_agents_data.pkl"]
+  L --> M["Append to prompt<br/>for consistent improvements"]
+```
+
+### During training runs
+
+- On each iteration, the system records for every agent:
+  - `initial_output`: the agent’s first answer
+  - `human_feedback`: your inline feedback when prompted
+  - `improved_output`: the agent’s follow-up answer after feedback
+- This data is stored in a working file named `training_data.pkl` keyed by the agent’s internal ID and iteration.
+- While training is active, the agent automatically appends your prior human feedback to its prompt to enforce those instructions on subsequent attempts within the training session.
+  Training is interactive: tasks set `human_input = true`, so running in a non-interactive environment will block on user input.
+
+### After training completes
+
+- When `train(...)` finishes, CrewAI evaluates the collected training data per agent and produces a consolidated result containing:
+  - `suggestions`: clear, actionable instructions distilled from your feedback and the difference between initial/improved outputs
+  - `quality`: a 0–10 score capturing improvement
+  - `final_summary`: a step-by-step set of action items for future tasks
+- These consolidated results are saved to the filename you pass to `train(...)` (default via CLI is `trained_agents_data.pkl`). Entries are keyed by the agent’s `role` so they can be applied across sessions.
+- During normal (non-training) execution, each agent automatically loads its consolidated `suggestions` and appends them to the task prompt as mandatory instructions. This gives you consistent improvements without changing your agent definitions.
+
+### File summary
+
+- `training_data.pkl` (ephemeral, per-session):
+  - Structure: `agent_id -> { iteration_number: { initial_output, human_feedback, improved_output } }`
+  - Purpose: capture raw data and human feedback during training
+  - Location: saved in the current working directory (CWD)
+- `trained_agents_data.pkl` (or your custom filename):
+  - Structure: `agent_role -> { suggestions: string[], quality: number, final_summary: string }`
+  - Purpose: persist consolidated guidance for future runs
+  - Location: written to the CWD by default; use `-f` to set a custom (including absolute) path
 
 ## Small Language Model Considerations
 
@@ -129,3 +179,18 @@ Happy training with CrewAI! 🚀
     </Warning>
   </Tab>
 </Tabs>
+
+### Key Points to Note
+
+- **Positive Integer Requirement:** Ensure that the number of iterations (`n_iterations`) is a positive integer. The code will raise a `ValueError` if this condition is not met.
+- **Filename Requirement:** Ensure that the filename ends with `.pkl`. The code will raise a `ValueError` if this condition is not met.
+- **Error Handling:** The code handles subprocess errors and unexpected exceptions, providing error messages to the user.
+- Trained guidance is applied at prompt time; it does not modify your Python/YAML agent configuration.
+- Agents automatically load trained suggestions from a file named `trained_agents_data.pkl` located in the current working directory. If you trained to a different filename, either rename it to `trained_agents_data.pkl` before running, or adjust the loader in code.
+- You can change the output filename when calling `crewai train` with `-f/--filename`. Absolute paths are supported if you want to save outside the CWD.
+
+It is important to note that the training process may take some time, depending on the complexity of your agents and will also require your feedback on each iteration.
+
+Once the training is complete, your agents will be equipped with enhanced capabilities and knowledge, ready to tackle complex tasks and provide more consistent and valuable insights.
+
+Remember to regularly update and retrain your agents to ensure they stay up-to-date with the latest information and advancements in the field.

docs/en/tools/overview.mdx

@@ -117,4 +117,19 @@ agent = Agent(
 )
 ```
 
+## **Max Usage Count**
+
+You can set a maximum usage count for a tool to prevent it from being used more than a certain number of times.
+By default, the max usage count is unlimited.
+
+
+
+```python
+from crewai_tools import FileReadTool
+
+tool = FileReadTool(max_usage_count=5, ...)
+```
+
+
+
 Ready to explore? Pick a category above to discover tools that fit your use case!

docs/ko/api-reference/inputs.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /inputs"
+description: "크루가 필요로 하는 입력 확인"
+openapi: "/enterprise-api.ko.yaml GET /inputs"
+---
+
+

docs/ko/api-reference/kickoff.mdx

@@ -0,0 +1,7 @@
+---
+title: "POST /kickoff"
+description: "크루 실행 시작"
+openapi: "/enterprise-api.ko.yaml POST /kickoff"
+---
+
+

docs/ko/api-reference/status.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /status/{kickoff_id}"
+description: "실행 상태 조회"
+openapi: "/enterprise-api.ko.yaml GET /status/{kickoff_id}"
+---
+
+

docs/pt-BR/api-reference/inputs.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /inputs"
+description: "Obter entradas necessárias para sua crew"
+openapi: "/enterprise-api.pt-BR.yaml GET /inputs"
+---
+
+

docs/pt-BR/api-reference/kickoff.mdx

@@ -0,0 +1,7 @@
+---
+title: "POST /kickoff"
+description: "Iniciar a execução da crew"
+openapi: "/enterprise-api.pt-BR.yaml POST /kickoff"
+---
+
+

docs/pt-BR/api-reference/status.mdx

@@ -0,0 +1,7 @@
+---
+title: "GET /status/{kickoff_id}"
+description: "Obter o status da execução"
+openapi: "/enterprise-api.pt-BR.yaml GET /status/{kickoff_id}"
+---
+
+

pyproject.toml

@@ -10,7 +10,7 @@ authors = [
 dependencies = [
     # Core Dependencies
     "pydantic>=2.4.2",
-    "openai<1.100.0",  # TODO: Temporarily pin due to https://github.com/openai/openai-python/issues/2564. Can revert once the issue is fixed.
+    "openai>=1.13.3",
     "litellm==1.74.9",
     "instructor>=1.3.3",
     # Text Processing

src/crewai/cli/update_crew.py

@@ -44,8 +44,9 @@ def migrate_pyproject(input_file, output_file):
         ]
         new_pyproject["project"]["requires-python"] = poetry_data.get("python")
     else:
-        # If it's already in the new format, just copy the project section
+        # If it's already in the new format, just copy the project and tool sections
         new_pyproject["project"] = pyproject_data.get("project", {})
+        new_pyproject["tool"] = pyproject_data.get("tool", {})
 
     # Migrate or copy dependencies
     if "dependencies" in new_pyproject["project"]:

src/crewai/flow/flow.py

@@ -913,17 +913,52 @@ class Flow(Generic[T], metaclass=FlowMeta):
         - Triggers execution of any listeners waiting on this start method
         - Part of the flow's initialization sequence
         - Skips execution if method was already completed (e.g., after reload)
+        - Automatically injects crewai_trigger_payload if available in flow inputs
         """
         if start_method_name in self._completed_methods:
             last_output = self._method_outputs[-1] if self._method_outputs else None
             await self._execute_listeners(start_method_name, last_output)
             return
 
+        method = self._methods[start_method_name]
+        enhanced_method = self._inject_trigger_payload_for_start_method(method)
+
         result = await self._execute_method(
-            start_method_name, self._methods[start_method_name]
+            start_method_name, enhanced_method
         )
         await self._execute_listeners(start_method_name, result)
 
+    def _inject_trigger_payload_for_start_method(self, original_method: Callable) -> Callable:
+        def prepare_kwargs(*args, **kwargs):
+            inputs = baggage.get_baggage("flow_inputs") or {}
+            trigger_payload = inputs.get("crewai_trigger_payload")
+
+            sig = inspect.signature(original_method)
+            accepts_trigger_payload = "crewai_trigger_payload" in sig.parameters
+
+            if trigger_payload is not None and accepts_trigger_payload:
+                kwargs["crewai_trigger_payload"] = trigger_payload
+            elif trigger_payload is not None:
+                self._log_flow_event(
+                    f"Trigger payload available but {original_method.__name__} doesn't accept crewai_trigger_payload parameter",
+                    color="yellow"
+                )
+            return args, kwargs
+
+        if asyncio.iscoroutinefunction(original_method):
+            async def enhanced_method(*args, **kwargs):
+                args, kwargs = prepare_kwargs(*args, **kwargs)
+                return await original_method(*args, **kwargs)
+        else:
+            def enhanced_method(*args, **kwargs):
+                args, kwargs = prepare_kwargs(*args, **kwargs)
+                return original_method(*args, **kwargs)
+
+        enhanced_method.__name__ = original_method.__name__
+        enhanced_method.__doc__ = original_method.__doc__
+
+        return enhanced_method
+
     async def _execute_method(
         self, method_name: str, method: Callable, *args: Any, **kwargs: Any
     ) -> Any:

src/crewai/rag/core/__init__.py

@@ -0,0 +1 @@
+"""Core abstract base classes and protocols for RAG  systems."""

src/crewai/rag/core/base_client.py

@@ -0,0 +1,433 @@
+"""Protocol for vector database client implementations."""
+
+from abc import abstractmethod
+from typing import Any, Protocol, runtime_checkable, TypedDict, Annotated
+from typing_extensions import Unpack, Required
+
+
+from crewai.rag.types import (
+    EmbeddingFunction,
+    BaseRecord,
+    SearchResult,
+)
+
+
+class BaseCollectionParams(TypedDict):
+    """Base parameters for collection operations.
+
+    Attributes:
+        collection_name: The name of the collection/index to operate on.
+    """
+
+    collection_name: Required[
+        Annotated[
+            str,
+            "Name of the collection/index. Implementations may have specific constraints (e.g., character limits, allowed characters, case sensitivity).",
+        ]
+    ]
+
+
+class BaseCollectionAddParams(BaseCollectionParams):
+    """Parameters for adding documents to a collection.
+
+    Extends BaseCollectionParams with document-specific fields.
+
+    Attributes:
+        collection_name: The name of the collection to add documents to.
+        documents: List of BaseRecord dictionaries containing document data.
+    """
+
+    documents: list[BaseRecord]
+
+
+class BaseCollectionSearchParams(BaseCollectionParams, total=False):
+    """Parameters for searching within a collection.
+
+    Extends BaseCollectionParams with search-specific optional fields.
+    All fields except collection_name and query are optional.
+
+    Attributes:
+        query: The text query to search for (required).
+        limit: Maximum number of results to return.
+        metadata_filter: Filter results by metadata fields.
+        score_threshold: Minimum similarity score for results (0-1).
+    """
+
+    query: Required[str]
+    limit: int
+    metadata_filter: dict[str, Any]
+    score_threshold: float
+
+
+@runtime_checkable
+class BaseClient(Protocol):
+    """Protocol for vector store client implementations.
+
+    This protocol defines the interface that all vector store client implementations
+    must follow. It provides a consistent API for storing and retrieving
+    documents with their vector embeddings across different vector database
+    backends (e.g., Qdrant, ChromaDB, Weaviate). Implementing classes should
+    handle connection management, data persistence, and vector similarity
+    search operations specific to their backend.
+
+    Implementation Guidelines:
+        Implementations should accept BaseClientParams in their constructor to allow
+        passing pre-configured client instances:
+
+        class MyVectorClient:
+            def __init__(self, client: Any | None = None, **kwargs):
+                if client:
+                    self.client = client
+                else:
+                    self.client = self._create_default_client(**kwargs)
+
+    Notes:
+      This protocol replaces the former BaseRAGStorage abstraction,
+      providing a cleaner interface for vector store operations.
+
+    Attributes:
+        embedding_function: Callable that takes a list of text strings
+            and returns a list of embedding vectors. Implementations
+            should always provide a default embedding function.
+        client: The underlying vector database client instance. This could be
+            passed via BaseClientParams during initialization or created internally.
+    """
+
+    client: Any
+    embedding_function: EmbeddingFunction
+
+    @abstractmethod
+    def create_collection(self, **kwargs: Unpack[BaseCollectionParams]) -> None:
+        """Create a new collection/index in the vector database.
+
+        Keyword Args:
+            collection_name: The name of the collection to create. Must be unique within
+                the vector database instance.
+
+        Raises:
+            ValueError: If collection name already exists.
+            ConnectionError: If unable to connect to the vector database backend.
+        """
+        ...
+
+    @abstractmethod
+    async def acreate_collection(self, **kwargs: Unpack[BaseCollectionParams]) -> None:
+        """Create a new collection/index in the vector database asynchronously.
+
+        Keyword Args:
+            collection_name: The name of the collection to create. Must be unique within
+                the vector database instance.
+
+        Raises:
+            ValueError: If collection name already exists.
+            ConnectionError: If unable to connect to the vector database backend.
+        """
+        ...
+
+    @abstractmethod
+    def get_or_create_collection(self, **kwargs: Unpack[BaseCollectionParams]) -> Any:
+        """Get an existing collection or create it if it doesn't exist.
+
+        This method provides a convenient way to ensure a collection exists
+        without having to check for its existence first.
+
+        Keyword Args:
+            collection_name: The name of the collection to get or create.
+
+        Returns:
+            A collection object whose type depends on the backend implementation.
+            This could be a collection reference, ID, or client object.
+
+        Raises:
+            ValueError: If unable to create the collection.
+            ConnectionError: If unable to connect to the vector database backend.
+        """
+        ...
+
+    @abstractmethod
+    async def aget_or_create_collection(
+        self, **kwargs: Unpack[BaseCollectionParams]
+    ) -> Any:
+        """Get an existing collection or create it if it doesn't exist asynchronously.
+
+        Keyword Args:
+            collection_name: The name of the collection to get or create.
+
+        Returns:
+            A collection object whose type depends on the backend implementation.
+
+        Raises:
+            ValueError: If unable to create the collection.
+            ConnectionError: If unable to connect to the vector database backend.
+        """
+        ...
+
+    @abstractmethod
+    def add_documents(self, **kwargs: Unpack[BaseCollectionAddParams]) -> None:
+        """Add documents with their embeddings to a collection.
+
+        This method performs an upsert operation - if a document with the same ID
+        already exists, it will be updated with the new content and metadata.
+
+        Implementations should handle embedding generation internally based on
+        the configured embedding function.
+
+        Keyword Args:
+            collection_name: The name of the collection to add documents to.
+            documents: List of BaseRecord dicts containing:
+                - content: The text content (required)
+                - doc_id: Optional unique identifier (auto-generated from content hash if missing)
+                - metadata: Optional metadata dictionary
+                Embeddings will be generated automatically.
+
+        Raises:
+            ValueError: If collection doesn't exist or documents list is empty.
+            TypeError: If documents are not BaseRecord dict instances.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>> from crewai.rag.types import BaseRecord
+            >>> client = ChromaDBClient()
+            >>>
+            >>> records: list[BaseRecord] = [
+            ...     {
+            ...         "content": "Machine learning basics",
+            ...         "metadata": {"source": "file3", "topic": "ML"}
+            ...     },
+            ...     {
+            ...         "doc_id": "custom_id",
+            ...         "content": "Deep learning fundamentals",
+            ...         "metadata": {"source": "file4", "topic": "DL"}
+            ...     }
+            ... ]
+            >>> client.add_documents(collection_name="my_docs", documents=records)
+            >>>
+            >>> records_with_id: list[BaseRecord] = [
+            ...     {
+            ...         "doc_id": "nlp_001",
+            ...         "content": "Advanced NLP techniques",
+            ...         "metadata": {"source": "file5", "topic": "NLP"}
+            ...     }
+            ... ]
+            >>> client.add_documents(collection_name="my_docs", documents=records_with_id)
+        """
+        ...
+
+    @abstractmethod
+    async def aadd_documents(self, **kwargs: Unpack[BaseCollectionAddParams]) -> None:
+        """Add documents with their embeddings to a collection asynchronously.
+
+        Implementations should handle embedding generation internally based on
+        the configured embedding function.
+
+        Keyword Args:
+            collection_name: The name of the collection to add documents to.
+            documents: List of BaseRecord dicts containing:
+                - content: The text content (required)
+                - doc_id: Optional unique identifier (auto-generated from content hash if missing)
+                - metadata: Optional metadata dictionary
+                Embeddings will be generated automatically.
+
+        Raises:
+            ValueError: If collection doesn't exist or documents list is empty.
+            TypeError: If documents are not BaseRecord dict instances.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> import asyncio
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>> from crewai.rag.types import BaseRecord
+            >>>
+            >>> async def add_documents():
+            ...     client = ChromaDBClient()
+            ...
+            ...     records: list[BaseRecord] = [
+            ...         {
+            ...             "doc_id": "doc2",
+            ...             "content": "Async operations in Python",
+            ...             "metadata": {"source": "file2", "topic": "async"}
+            ...         }
+            ...     ]
+            ...     await client.aadd_documents(collection_name="my_docs", documents=records)
+            ...
+            >>> asyncio.run(add_documents())
+        """
+        ...
+
+    @abstractmethod
+    def search(
+        self, **kwargs: Unpack[BaseCollectionSearchParams]
+    ) -> list[SearchResult]:
+        """Search for similar documents using a query.
+
+        Performs a vector similarity search to find the most similar documents
+        to the provided query.
+
+        Keyword Args:
+            collection_name: The name of the collection to search in.
+            query: The text query to search for. The implementation handles
+                embedding generation internally.
+            limit: Maximum number of results to return. Defaults to 10.
+            metadata_filter: Optional metadata filter to apply to the search. The exact
+                format depends on the backend, but typically supports equality
+                and range queries on metadata fields.
+            score_threshold: Optional minimum similarity score threshold. Only
+                results with scores >= this threshold will be returned. The
+                score interpretation depends on the distance metric used.
+
+        Returns:
+            A list of SearchResult dictionaries ordered by similarity score in
+            descending order. Each result contains:
+                - id: Document ID
+                - content: Document text content
+                - metadata: Document metadata
+                - score: Similarity score (0-1, higher is better)
+
+        Raises:
+            ValueError: If collection doesn't exist.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>> client = ChromaDBClient()
+            >>>
+            >>> results = client.search(
+            ...     collection_name="my_docs",
+            ...     query="What is machine learning?",
+            ...     limit=5,
+            ...     metadata_filter={"source": "file1"},
+            ...     score_threshold=0.7
+            ... )
+            >>> for result in results:
+            ...     print(f"{result['id']}: {result['score']:.2f}")
+        """
+        ...
+
+    @abstractmethod
+    async def asearch(
+        self, **kwargs: Unpack[BaseCollectionSearchParams]
+    ) -> list[SearchResult]:
+        """Search for similar documents using a query asynchronously.
+
+        Keyword Args:
+            collection_name: The name of the collection to search in.
+            query: The text query to search for. The implementation handles
+                embedding generation internally.
+            limit: Maximum number of results to return. Defaults to 10.
+            metadata_filter: Optional metadata filter to apply to the search.
+            score_threshold: Optional minimum similarity score threshold.
+
+        Returns:
+            A list of SearchResult dictionaries ordered by similarity score.
+
+        Raises:
+            ValueError: If collection doesn't exist.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> import asyncio
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>>
+            >>> async def search_documents():
+            ...     client = ChromaDBClient()
+            ...     results = await client.asearch(
+            ...         collection_name="my_docs",
+            ...         query="Python programming best practices",
+            ...         limit=5,
+            ...         metadata_filter={"source": "file1"},
+            ...         score_threshold=0.7
+            ...     )
+            ...     for result in results:
+            ...         print(f"{result['id']}: {result['score']:.2f}")
+            ...
+            >>> asyncio.run(search_documents())
+        """
+        ...
+
+    @abstractmethod
+    def delete_collection(self, **kwargs: Unpack[BaseCollectionParams]) -> None:
+        """Delete a collection and all its data.
+
+        This operation is irreversible and will permanently remove all documents,
+        embeddings, and metadata associated with the collection.
+
+        Keyword Args:
+            collection_name: The name of the collection to delete.
+
+        Raises:
+            ValueError: If the collection doesn't exist.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>> client = ChromaDBClient()
+            >>> client.delete_collection(collection_name="old_docs")
+            >>> print("Collection 'old_docs' deleted successfully")
+        """
+        ...
+
+    @abstractmethod
+    async def adelete_collection(self, **kwargs: Unpack[BaseCollectionParams]) -> None:
+        """Delete a collection and all its data asynchronously.
+
+        Keyword Args:
+            collection_name: The name of the collection to delete.
+
+        Raises:
+            ValueError: If the collection doesn't exist.
+            ConnectionError: If unable to connect to the vector database backend.
+
+        Example:
+            >>> import asyncio
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>>
+            >>> async def delete_old_collection():
+            ...     client = ChromaDBClient()
+            ...     await client.adelete_collection(collection_name="old_docs")
+            ...     print("Collection 'old_docs' deleted successfully")
+            ...
+            >>> asyncio.run(delete_old_collection())
+        """
+        ...
+
+    @abstractmethod
+    def reset(self) -> None:
+        """Reset the vector database by deleting all collections and data.
+
+        This method provides a way to completely clear the vector database,
+        removing all collections and their contents. Use with caution as
+        this operation is irreversible.
+
+        Raises:
+            ConnectionError: If unable to connect to the vector database backend.
+            PermissionError: If the operation is not allowed by the backend.
+
+        Example:
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>> client = ChromaDBClient()
+            >>> client.reset()
+            >>> print("Vector database completely reset - all data deleted")
+        """
+        ...
+
+    @abstractmethod
+    async def areset(self) -> None:
+        """Reset the vector database by deleting all collections and data asynchronously.
+
+        Raises:
+            ConnectionError: If unable to connect to the vector database backend.
+            PermissionError: If the operation is not allowed by the backend.
+
+        Example:
+            >>> import asyncio
+            >>> from crewai.rag.chromadb.client import ChromaDBClient
+            >>>
+            >>> async def reset_database():
+            ...     client = ChromaDBClient()
+            ...     await client.areset()
+            ...     print("Vector database completely reset - all data deleted")
+            ...
+            >>> asyncio.run(reset_database())
+        """
+        ...

src/crewai/rag/core/base_provider.py

@@ -0,0 +1,30 @@
+"""Base provider protocol for vector database client creation."""
+
+from abc import ABC
+from typing import Any, Protocol, runtime_checkable, Union
+from pydantic import BaseModel, Field
+
+from crewai.rag.types import EmbeddingFunction
+from crewai.rag.embeddings.types import EmbeddingOptions
+
+
+class BaseProviderOptions(BaseModel, ABC):
+    """Base configuration for all provider options."""
+
+    client_type: str = Field(..., description="Type of client to create")
+    embedding_config: Union[EmbeddingOptions, EmbeddingFunction, None] = Field(
+        default=None,
+        description="Embedding configuration - either options for built-in providers or a custom function",
+    )
+    options: Any = Field(
+        default=None, description="Additional provider-specific options"
+    )
+
+
+@runtime_checkable
+class BaseProvider(Protocol):
+    """Protocol for vector database client providers."""
+
+    def __call__(self, options: BaseProviderOptions) -> Any:
+        """Create and return a configured client instance."""
+        ...

src/crewai/rag/embeddings/factory.py

@@ -0,0 +1,148 @@
+"""Minimal embedding function factory for CrewAI."""
+
+import os
+
+from chromadb import EmbeddingFunction
+from chromadb.utils.embedding_functions.amazon_bedrock_embedding_function import (
+    AmazonBedrockEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.cohere_embedding_function import (
+    CohereEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.google_embedding_function import (
+    GooglePalmEmbeddingFunction,
+    GoogleGenerativeAiEmbeddingFunction,
+    GoogleVertexEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.huggingface_embedding_function import (
+    HuggingFaceEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.instructor_embedding_function import (
+    InstructorEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.jina_embedding_function import (
+    JinaEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.ollama_embedding_function import (
+    OllamaEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.onnx_mini_lm_l6_v2 import ONNXMiniLM_L6_V2
+from chromadb.utils.embedding_functions.open_clip_embedding_function import (
+    OpenCLIPEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.openai_embedding_function import (
+    OpenAIEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.roboflow_embedding_function import (
+    RoboflowEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.sentence_transformer_embedding_function import (
+    SentenceTransformerEmbeddingFunction,
+)
+from chromadb.utils.embedding_functions.text2vec_embedding_function import (
+    Text2VecEmbeddingFunction,
+)
+
+from crewai.rag.embeddings.types import EmbeddingOptions
+
+
+def get_embedding_function(
+    config: EmbeddingOptions | dict | None = None,
+) -> EmbeddingFunction:
+    """Get embedding function - delegates to ChromaDB.
+
+    Args:
+        config: Optional configuration - either an EmbeddingOptions object or a dict with:
+            - provider: The embedding provider to use (default: "openai")
+            - Any other provider-specific parameters
+
+    Returns:
+        EmbeddingFunction instance ready for use with ChromaDB
+
+    Supported providers:
+        - openai: OpenAI embeddings (default)
+        - cohere: Cohere embeddings
+        - ollama: Ollama local embeddings
+        - huggingface: HuggingFace embeddings
+        - sentence-transformer: Local sentence transformers
+        - instructor: Instructor embeddings for specialized tasks
+        - google-palm: Google PaLM embeddings
+        - google-generativeai: Google Generative AI embeddings
+        - google-vertex: Google Vertex AI embeddings
+        - amazon-bedrock: AWS Bedrock embeddings
+        - jina: Jina AI embeddings
+        - roboflow: Roboflow embeddings for vision tasks
+        - openclip: OpenCLIP embeddings for multimodal tasks
+        - text2vec: Text2Vec embeddings
+        - onnx: ONNX MiniLM-L6-v2 (no API key needed, included with ChromaDB)
+
+    Examples:
+        # Use default OpenAI with retry logic
+        >>> embedder = get_embedding_function()
+
+        # Use Cohere with dict
+        >>> embedder = get_embedding_function({
+        ...     "provider": "cohere",
+        ...     "api_key": "your-key",
+        ...     "model_name": "embed-english-v3.0"
+        ... })
+
+        # Use with EmbeddingOptions
+        >>> embedder = get_embedding_function(
+        ...     EmbeddingOptions(provider="sentence-transformer", model_name="all-MiniLM-L6-v2")
+        ... )
+
+        # Use local sentence transformers (no API key needed)
+        >>> embedder = get_embedding_function({
+        ...     "provider": "sentence-transformer",
+        ...     "model_name": "all-MiniLM-L6-v2"
+        ... })
+
+        # Use Ollama for local embeddings
+        >>> embedder = get_embedding_function({
+        ...     "provider": "ollama",
+        ...     "model_name": "nomic-embed-text"
+        ... })
+
+        # Use ONNX (no API key needed)
+        >>> embedder = get_embedding_function({
+        ...     "provider": "onnx"
+        ... })
+    """
+    if config is None:
+        return OpenAIEmbeddingFunction(
+            api_key=os.getenv("OPENAI_API_KEY"), model_name="text-embedding-3-small"
+        )
+
+    # Handle EmbeddingOptions object
+    if isinstance(config, EmbeddingOptions):
+        config_dict = config.model_dump(exclude_none=True)
+    else:
+        config_dict = config.copy()
+
+    provider = config_dict.pop("provider", "openai")
+
+    embedding_functions = {
+        "openai": OpenAIEmbeddingFunction,
+        "cohere": CohereEmbeddingFunction,
+        "ollama": OllamaEmbeddingFunction,
+        "huggingface": HuggingFaceEmbeddingFunction,
+        "sentence-transformer": SentenceTransformerEmbeddingFunction,
+        "instructor": InstructorEmbeddingFunction,
+        "google-palm": GooglePalmEmbeddingFunction,
+        "google-generativeai": GoogleGenerativeAiEmbeddingFunction,
+        "google-vertex": GoogleVertexEmbeddingFunction,
+        "amazon-bedrock": AmazonBedrockEmbeddingFunction,
+        "jina": JinaEmbeddingFunction,
+        "roboflow": RoboflowEmbeddingFunction,
+        "openclip": OpenCLIPEmbeddingFunction,
+        "text2vec": Text2VecEmbeddingFunction,
+        "onnx": ONNXMiniLM_L6_V2,
+    }
+
+    if provider not in embedding_functions:
+        raise ValueError(
+            f"Unsupported provider: {provider}. "
+            f"Available providers: {list(embedding_functions.keys())}"
+        )
+    return embedding_functions[provider](**config_dict)

src/crewai/rag/embeddings/types.py

@@ -0,0 +1,62 @@
+"""Type definitions for the embeddings module."""
+
+from typing import Literal
+from pydantic import BaseModel, Field, SecretStr
+
+from crewai.rag.types import EmbeddingFunction
+
+
+EmbeddingProvider = Literal[
+    "openai",
+    "cohere",
+    "ollama",
+    "huggingface",
+    "sentence-transformer",
+    "instructor",
+    "google-palm",
+    "google-generativeai",
+    "google-vertex",
+    "amazon-bedrock",
+    "jina",
+    "roboflow",
+    "openclip",
+    "text2vec",
+    "onnx",
+]
+"""Supported embedding providers.
+
+These correspond to the embedding functions available in ChromaDB's
+embedding_functions module. Each provider has specific requirements
+and configuration options.
+"""
+
+
+class EmbeddingOptions(BaseModel):
+    """Configuration options for embedding providers.
+
+    Generic attributes that can be passed to get_embedding_function
+    to configure various embedding providers.
+    """
+
+    provider: EmbeddingProvider = Field(
+        ..., description="Embedding provider name (e.g., 'openai', 'cohere', 'onnx')"
+    )
+    model_name: str | None = Field(
+        default=None, description="Model name for the embedding provider"
+    )
+    api_key: SecretStr | None = Field(
+        default=None, description="API key for the embedding provider"
+    )
+
+
+class EmbeddingConfig(BaseModel):
+    """Configuration wrapper for embedding functions.
+
+    Accepts either a pre-configured EmbeddingFunction or EmbeddingOptions
+    to create one. This provides flexibility in how embeddings are configured.
+
+    Attributes:
+        function: Either a callable EmbeddingFunction or EmbeddingOptions to create one
+    """
+
+    function: EmbeddingFunction | EmbeddingOptions

src/crewai/rag/types.py

@@ -0,0 +1,50 @@
+"""Type definitions for RAG (Retrieval-Augmented Generation) systems."""
+
+from collections.abc import Callable, Mapping
+from typing import TypeAlias, TypedDict, Any
+
+from typing_extensions import Required
+
+
+class BaseRecord(TypedDict, total=False):
+    """A typed dictionary representing a document record.
+
+    Attributes:
+        doc_id: Optional unique identifier for the document. If not provided,
+            a content-based ID will be generated using SHA256 hash.
+        content: The text content of the document (required)
+        metadata: Optional metadata associated with the document
+    """
+
+    doc_id: str
+    content: Required[str]
+    metadata: (
+        Mapping[str, str | int | float | bool]
+        | list[Mapping[str, str | int | float | bool]]
+    )
+
+
+DenseVector: TypeAlias = list[float]
+IntVector: TypeAlias = list[int]
+
+EmbeddingFunction: TypeAlias = Callable[..., Any]
+
+
+class SearchResult(TypedDict):
+    """Standard search result format for vector store queries.
+
+    This provides a consistent interface for search results across different
+    vector store implementations. Each implementation should convert their
+    native result format to this standard format.
+
+    Attributes:
+        id: Unique identifier of the document
+        content: The text content of the document
+        metadata: Optional metadata associated with the document
+        score: Similarity score (higher is better, typically between 0 and 1)
+    """
+
+    id: str
+    content: str
+    metadata: dict[str, Any]
+    score: float

src/crewai/tools/base_tool.py

@@ -14,12 +14,14 @@ from pydantic import BaseModel as PydanticBaseModel
 
 from crewai.tools.structured_tool import CrewStructuredTool
 
+
 class EnvVar(BaseModel):
     name: str
     description: str
     required: bool = True
     default: Optional[str] = None
 
+
 class BaseTool(BaseModel, ABC):
     class _ArgsSchemaPlaceholder(PydanticBaseModel):
         pass
@@ -108,7 +110,7 @@ class BaseTool(BaseModel, ABC):
     def to_structured_tool(self) -> CrewStructuredTool:
         """Convert this tool to a CrewStructuredTool instance."""
         self._set_args_schema()
-        return CrewStructuredTool(
+        structured_tool = CrewStructuredTool(
             name=self.name,
             description=self.description,
             args_schema=self.args_schema,
@@ -117,6 +119,8 @@ class BaseTool(BaseModel, ABC):
             max_usage_count=self.max_usage_count,
             current_usage_count=self.current_usage_count,
         )
+        structured_tool._original_tool = self
+        return structured_tool
 
     @classmethod
     def from_langchain(cls, tool: Any) -> "BaseTool":
@@ -276,7 +280,9 @@ def to_langchain(
     return [t.to_structured_tool() if isinstance(t, BaseTool) else t for t in tools]
 
 
-def tool(*args, result_as_answer: bool = False, max_usage_count: int | None = None) -> Callable:
+def tool(
+    *args, result_as_answer: bool = False, max_usage_count: int | None = None
+) -> Callable:
     """
     Decorator to create a tool from a function.
 

src/crewai/tools/structured_tool.py

@@ -10,6 +10,11 @@ from pydantic import BaseModel, Field, create_model
 
 from crewai.utilities.logger import Logger
 
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from crewai.tools.base_tool import BaseTool
+
 
 class CrewStructuredTool:
     """A structured tool that can operate on any number of inputs.
@@ -18,6 +23,8 @@ class CrewStructuredTool:
     that integrates better with CrewAI's ecosystem.
     """
 
+    _original_tool: BaseTool | None = None
+
     def __init__(
         self,
         name: str,
@@ -47,6 +54,7 @@ class CrewStructuredTool:
         self.result_as_answer = result_as_answer
         self.max_usage_count = max_usage_count
         self.current_usage_count = current_usage_count
+        self._original_tool = None
 
         # Validate the function signature matches the schema
         self._validate_function_signature()
@@ -219,6 +227,8 @@ class CrewStructuredTool:
         """
         parsed_args = self._parse_args(input)
 
+        self._increment_usage_count()
+
         if inspect.iscoroutinefunction(self.func):
             return await self.func(**parsed_args, **kwargs)
         else:
@@ -242,6 +252,8 @@ class CrewStructuredTool:
         """Main method for tool execution."""
         parsed_args = self._parse_args(input)
 
+        self._increment_usage_count()
+
         if inspect.iscoroutinefunction(self.func):
             result = asyncio.run(self.func(**parsed_args, **kwargs))
             return result
@@ -253,6 +265,12 @@ class CrewStructuredTool:
 
         return result
 
+    def _increment_usage_count(self) -> None:
+        """Increment the usage count."""
+        self.current_usage_count += 1
+        if self._original_tool is not None:
+            self._original_tool.current_usage_count = self.current_usage_count
+
     @property
     def args(self) -> dict:
         """Get the tool's input arguments schema."""

src/crewai/utilities/events/event_listener.py

@@ -161,8 +161,10 @@ class EventListener(BaseEventListener):
         def on_task_started(source, event: TaskStartedEvent):
             span = self._telemetry.task_started(crew=source.agent.crew, task=source)
             self.execution_spans[source] = span
+            # Pass both task ID and task name (if set)
+            task_name = source.name if hasattr(source, 'name') and source.name else None
             self.formatter.create_task_branch(
-                self.formatter.current_crew_tree, source.id
+                self.formatter.current_crew_tree, source.id, task_name
             )
 
         @crewai_event_bus.on(TaskCompletedEvent)
@@ -173,11 +175,14 @@ class EventListener(BaseEventListener):
                 self._telemetry.task_ended(span, source, source.agent.crew)
             self.execution_spans[source] = None
 
+            # Pass task name if it exists
+            task_name = source.name if hasattr(source, 'name') and source.name else None
             self.formatter.update_task_status(
                 self.formatter.current_crew_tree,
                 source.id,
                 source.agent.role,
                 "completed",
+                task_name
             )
 
         @crewai_event_bus.on(TaskFailedEvent)
@@ -188,11 +193,14 @@ class EventListener(BaseEventListener):
                     self._telemetry.task_ended(span, source, source.agent.crew)
                 self.execution_spans[source] = None
 
+            # Pass task name if it exists
+            task_name = source.name if hasattr(source, 'name') and source.name else None
             self.formatter.update_task_status(
                 self.formatter.current_crew_tree,
                 source.id,
                 source.agent.role,
                 "failed",
+                task_name
             )
 
         # ----------- AGENT EVENTS -----------

src/crewai/utilities/events/utils/console_formatter.py

@@ -220,14 +220,22 @@ class ConsoleFormatter:
         return tree
 
     def create_task_branch(
-        self, crew_tree: Optional[Tree], task_id: str
+        self, crew_tree: Optional[Tree], task_id: str, task_name: Optional[str] = None
     ) -> Optional[Tree]:
         """Create and initialize a task branch."""
         if not self.verbose:
             return None
 
         task_content = Text()
-        task_content.append(f"📋 Task: {task_id}", style="yellow bold")
+        
+        # Display task name if available, otherwise just the ID
+        if task_name:
+            task_content.append("📋 Task: ", style="yellow bold")
+            task_content.append(f"{task_name}", style="yellow bold")
+            task_content.append(f" (ID: {task_id})", style="yellow dim")
+        else:
+            task_content.append(f"📋 Task: {task_id}", style="yellow bold")
+            
         task_content.append("\nStatus: ", style="white")
         task_content.append("Executing Task...", style="yellow dim")
 
@@ -251,6 +259,7 @@ class ConsoleFormatter:
         task_id: str,
         agent_role: str,
         status: str = "completed",
+        task_name: Optional[str] = None,
     ) -> None:
         """Update task status in the tree."""
         if not self.verbose or crew_tree is None:
@@ -270,8 +279,13 @@ class ConsoleFormatter:
             if str(task_id) in str(branch.label):
                 # Build label without introducing stray blank lines
                 task_content = Text()
-                # First line: Task ID
-                task_content.append(f"📋 Task: {task_id}", style=f"{style} bold")
+                # First line: Task ID/name
+                if task_name:
+                    task_content.append("📋 Task: ", style=f"{style} bold")
+                    task_content.append(f"{task_name}", style=f"{style} bold")
+                    task_content.append(f" (ID: {task_id})", style=f"{style} dim")
+                else:
+                    task_content.append(f"📋 Task: {task_id}", style=f"{style} bold")
 
                 # Second line: Assigned to
                 task_content.append("\nAssigned to: ", style="white")
@@ -285,8 +299,9 @@ class ConsoleFormatter:
                 break
 
         # Show status panel
+        display_name = task_name if task_name else str(task_id)
         content = self.create_status_content(
-            f"Task {status.title()}", str(task_id), style, Agent=agent_role
+            f"Task {status.title()}", display_name, style, Agent=agent_role
         )
         self.print_panel(content, panel_title, style)
 

tests/cassettes/test_max_usage_count_is_respected.yaml

@@ -0,0 +1,845 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}], "model": "gpt-4o", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '1452'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAAwAAAP//hFPRjtMwEHzPV6z83J5ySVvu8nYCIZ1ACKSCEOSUOs4mMefYlr2Boqr/
+        jpy0TQ7uxIsV7exMZmftQwTAZMUyYKLlJDqrlq837zbXn/flx6/fVPIp/fI+3ZbaxVv568ObR7YI
+        DFP+QEFn1pUwnVVI0ugRFg45YVC9frVeb5I0WacD0JkKVaA1lpYrs0ziZLWMb5bx5kRsjRToWQbf
+        IwCAw3AGi7rCPcsgXpwrHXrPG2TZpQmAOaNChXHvpSeuiS0mUBhNqAfXu90u19vW9E1LGdyDRqyA
+        DAiuFFCLIAkdJ6kbIGPGUi2dJyDZ4VWu70SYNpv6itB3rsO9tj1lcMiZDF8F4Z5ylkHO3g4qJ5rR
+        OTvOLTqse89DQrpXagZwrQ0NjCGchxNyvMShTGOdKf1fVFZLLX1bOOTe6DC6J2PZgB4jgIch9v5J
+        ksw601kqyDzi8Lvk9nbUY9OiJzRdn0AyxNWsnq4Wz+gVFRKXys8WxwQXLVYTddoy7ytpZkA0m/pf
+        N89pj5OPG/qv/AQIgZawKqzDSoqnE09tDsM7eKntkvJgmHl0P6XAgiS6sIkKa96r8Yoy/9sTdkUt
+        dYPOOjne09oWq01Z1zHG4oZFx+gPAAAA//8DAHvlcCKwAwAA
+    headers:
+      CF-RAY:
+      - 971b3f72effa6897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:35 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        path=/; expires=Tue, 19-Aug-25 17:37:35 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '2564'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '2751'
+      x-ratelimit-limit-project-tokens:
+      - '30000000'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-project-tokens:
+      - '29999674'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999674'
+      x-ratelimit-reset-project-tokens:
+      - 0s
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_d654df1116aa42ca8ee7d10b4b424303
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}, {"role": "assistant", "content":
+      "```\nThought: I need to call the iterating tool the first time.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"First iteration\"}\nObservation: Iteration 0: First
+      iteration"}], "model": "gpt-4o", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '1673'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAA4RTTW/bMAy9+1cQOidFlg8v8G1LB6zosbvNhaNItK3NFjWJ7loE+e+D7CR2uw67
+        CAIf3xP5SB0TAGG0yECoWrJqXTPfpffp2tWqu93u7m+Xu+3Dl6/m96+X+pB+VmIWGXT4gYovrBtF
+        rWuQDdkBVh4lY1T98HGzSZer5SbtgZY0NpFWOZ6vab5cLNfzxXa+SM/EmozCIDL4ngAAHPszlmg1
+        PosMFrNLpMUQZIUiuyYBCE9NjAgZggksLYvZCCqyjLaver/f5/ZbTV1VcwZ3UMsnhHMXqIFrhNL4
+        wGAYvYyNgbQaLEaQwHlS52tMDajIaiCLN7n9pGJ6dmHaqmCi5hKHO+s6zuCYCxNvBeMz5yKDXDwM
+        KtcXc3GaVu+x7IKM5tmuaSaAtJa4Z/S+PZ6R09Wphirn6RDeUEVprAl14VEGstGVwOREj54SgMd+
+        It0rk4Xz1DoumH5i/9xqvRn0xLgDE3R7BplYNtP4avaOXqGRpWnCZKZCSVWjHqnjAshOG5oAyaTr
+        v6t5T3vofBjRf+VHQCl0jLpwHrVRrzse0zzGL/KvtKvLfcEioH8yCgs26OMkNJaya4btFeElMLZF
+        aWyF3nkzrHDpinV6KMsFLtRWJKfkDwAAAP//AwAMYddzywMAAA==
+    headers:
+      CF-RAY:
+      - 971b3f84cf146897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:38 GMT
+      Server:
+      - cloudflare
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1900'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '2551'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999629'
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_f4181fe581264993ac5c6deba4f1c287
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}, {"role": "assistant", "content":
+      "```\nThought: I need to call the iterating tool the first time.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"First iteration\"}\nObservation: Iteration 0: First
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the first iteration and need to proceed to the second one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Second iteration\"}\nObservation: Iteration 0: Second
+      iteration"}], "model": "gpt-4o", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '1922'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAAwAAAP//hFNNb9swDL37VxA6J0XmfCz1bRswNBh2GXrYsBSOItG2ElsUJLrIUOS/
+        D7KT2N067CIIfHxP5CP1kgAIo0UGQlWSVePq6afVl9Xq9HAI3z6uj8eHz6os9OGrvt98T9sfYhIZ
+        tD+g4ivrTlHjamRDtoeVR8kYVd+9Xy5X6TxdrjugIY11pJWOpwuaprN0MZ2tp7PVhViRURhEBj8T
+        AICX7owlWo0nkcFsco00GIIsUWS3JADhqY4RIUMwgaVlMRlARZbRdlXvdrutfayoLSvOYAOVfEa4
+        dIEauEIIqMhqMIxexs5AWg0WI0rgPKnLNeZyZbwGsni3tR9UzM6uRFvmTFRf47CxruUMXrbCxFvO
+        eOKtyGArHjuR23tbcR4X77Fog4ze2bauR4C0lrhjdLY9XZDzzaiaSudpH/6gisJYE6rcowxkoymB
+        yYkOPScAT91A2lceC+epcZwzHbF7bn6/6PXEsAIjdH0BmVjWQ3wxTydv6OUaWZo6jEYqlFQV6oE6
+        zF+22tAISEZd/13NW9p95/2E/is/AEqhY9S586iNet3xkOYx/pB/pd1c7goWAf2zUZizQR8nobGQ
+        bd0vrwi/AmOTF8aW6J03/QYXLl+s9kUxw5lai+Sc/AYAAP//AwDpY7tZygMAAA==
+    headers:
+      CF-RAY:
+      - 971b3f958e746897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:39 GMT
+      Server:
+      - cloudflare
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '890'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '906'
+      x-ratelimit-limit-project-tokens:
+      - '30000000'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-project-tokens:
+      - '29999577'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999577'
+      x-ratelimit-reset-project-tokens:
+      - 0s
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_b3632e88268747218e4cc4cc08d87bca
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}, {"role": "assistant", "content":
+      "```\nThought: I need to call the iterating tool the first time.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"First iteration\"}\nObservation: Iteration 0: First
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the first iteration and need to proceed to the second one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Second iteration\"}\nObservation: Iteration 0: Second
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the second iteration and need to proceed to the third one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Third iteration\"}\nObservation: Iteration 0: Third
+      iteration\n\n\nYou ONLY have access to the following tools, and should NEVER
+      make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool Arguments:
+      {''input_text'': {''description'': None, ''type'': ''str''}}\nTool Description:
+      A tool that iterates a given number of times\n\nIMPORTANT: Use the following
+      format in your response:\n\n```\nThought: you should always think about what
+      to do\nAction: the action to take, only one name of [iterating_tool], just the
+      name, exactly as it''s written.\nAction Input: the input to the action, just
+      a simple JSON object, enclosed in curly braces, using \" to wrap keys and values.\nObservation:
+      the result of the action\n```\n\nOnce all necessary information is gathered,
+      return the following format:\n\n```\nThought: I now know the final answer\nFinal
+      Answer: the final answer to the original input question\n```"}], "model": "gpt-4o",
+      "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '3018'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAAwAAAP//hFPBbtswDL37Kwidk8JzErfxbdgwtNh1GwrMhaPItK3MFgWJLloE+fdB
+        dhK7W4ddBIGP74l8pI4RgNClyECoRrLqbLv8lH5Nb8k9Hu7T5PP37f361R1Wj1us6IeVYhEYtD+g
+        4gvrRlFnW2RNZoSVQ8kYVD/cbjZpsko22wHoqMQ20GrLyzUtkzhZL+O7ZZyeiQ1phV5k8DMCADgO
+        ZyjRlPgiMogXl0iH3ssaRXZNAhCO2hAR0nvtWRoWiwlUZBjNUPVut8vNt4b6uuEMHqCRzwjnLrAE
+        bhC40a4EzehkaAykKcFgAAmsI3W+htSKescNkMGb3HxUIT27ME1dMFF7icODsT1ncMyFDreC8YVz
+        kUEuvowq1xdzcZpX77DqvQzmmb5tZ4A0hnhgDL49nZHT1amWauto7/+gikob7ZvCofRkgiueyYoB
+        PUUAT8NE+jcmC+uos1ww/cLhuTRJRz0x7cCEru7OIBPLdsZK14t39IoSWerWz2YqlFQNlhN1WgDZ
+        l5pmQDTr+u9q3tMeOx9H9F/5CVAKLWNZWIelVm87ntIchi/yr7Sry0PBwqN71goL1ujCJEqsZN+O
+        2yv8q2fsikqbGp11elzhyhbrdF9VMcbqTkSn6DcAAAD//wMACjMtRssDAAA=
+    headers:
+      CF-RAY:
+      - 971b3f9bbef86897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:40 GMT
+      Server:
+      - cloudflare
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1182'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '1208'
+      x-ratelimit-limit-project-tokens:
+      - '30000000'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-project-tokens:
+      - '29999320'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999320'
+      x-ratelimit-reset-project-tokens:
+      - 1ms
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 1ms
+      x-request-id:
+      - req_7fc641fabc634f29ae085ef176071402
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}, {"role": "assistant", "content":
+      "```\nThought: I need to call the iterating tool the first time.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"First iteration\"}\nObservation: Iteration 0: First
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the first iteration and need to proceed to the second one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Second iteration\"}\nObservation: Iteration 0: Second
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the second iteration and need to proceed to the third one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Third iteration\"}\nObservation: Iteration 0: Third
+      iteration\n\n\nYou ONLY have access to the following tools, and should NEVER
+      make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool Arguments:
+      {''input_text'': {''description'': None, ''type'': ''str''}}\nTool Description:
+      A tool that iterates a given number of times\n\nIMPORTANT: Use the following
+      format in your response:\n\n```\nThought: you should always think about what
+      to do\nAction: the action to take, only one name of [iterating_tool], just the
+      name, exactly as it''s written.\nAction Input: the input to the action, just
+      a simple JSON object, enclosed in curly braces, using \" to wrap keys and values.\nObservation:
+      the result of the action\n```\n\nOnce all necessary information is gathered,
+      return the following format:\n\n```\nThought: I now know the final answer\nFinal
+      Answer: the final answer to the original input question\n```"}, {"role": "assistant",
+      "content": "```\nThought: I have completed the third iteration and need to proceed
+      to the fourth one.\nAction: iterating_tool\nAction Input: {\"input_text\": \"Fourth
+      iteration\"}\nObservation: Iteration 0: Fourth iteration"}], "model": "gpt-4o",
+      "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '3267'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAA4RTTW/bMAy9+1cQOieFmyZO5ts2bECxDwztgB3mwlFk2tZmi4JEFx2C/PdBchKn
+        W4ddBIGP74l8pPYJgNCVyEGoVrLqbTd/m33INvJ+0NnnT/ff3tTZR72483eSv7xLScwCg3Y/UPGJ
+        daWotx2yJjPCyqFkDKrX69UqW9wssjQCPVXYBVpjeb6k+SJdLOfpZp5mR2JLWqEXOXxPAAD28Qwl
+        mgqfRA5RJkZ69F42KPJzEoBw1IWIkN5rz9KwmE2gIsNoYtXb7bYwX1sampZzuIVWPiIcu8AKuEWo
+        aXDcgmZ0MnQG0lRgMKAE1pE6XmOurrkFMnhVmNcqZOcnomlKJupOcbg1duAc9oXQ4VYyPnEhcijE
+        +yhyfq8Qh8viHdaDl8E7M3TdBSCNIY6MaNvDETmcjeqosY52/g+qqLXRvi0dSk8mmOKZrIjoIQF4
+        iAMZnnksrKPecsn0E+Nz2Xo16olpBSb05tURZGLZTfH19XL2gl5ZIUvd+YuRCiVVi9VEneYvh0rT
+        BZBcdP13NS9pj52PE/qv/AQohZaxKq3DSqvnHU9pDsMP+Vfa2eVYsPDoHrXCkjW6MIkKazl04/IK
+        /8sz9mWtTYPOOj1ucG3LZbar6xRTtRHJIfkNAAD//wMANy12ZMoDAAA=
+    headers:
+      CF-RAY:
+      - 971b3fa3ea2f6897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:41 GMT
+      Server:
+      - cloudflare
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '780'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '878'
+      x-ratelimit-limit-project-tokens:
+      - '30000000'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-project-tokens:
+      - '29999268'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999268'
+      x-ratelimit-reset-project-tokens:
+      - 1ms
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 1ms
+      x-request-id:
+      - req_603a6c645bac468888838d21c64db11f
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are Iterating Agent. You
+      are an agent that iterates a given number of times\nYour personal goal is: Call
+      the iterating tool 5 times\nYou ONLY have access to the following tools, and
+      should NEVER make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool
+      Arguments: {''input_text'': {''description'': None, ''type'': ''str''}}\nTool
+      Description: A tool that iterates a given number of times\n\nIMPORTANT: Use
+      the following format in your response:\n\n```\nThought: you should always think
+      about what to do\nAction: the action to take, only one name of [iterating_tool],
+      just the name, exactly as it''s written.\nAction Input: the input to the action,
+      just a simple JSON object, enclosed in curly braces, using \" to wrap keys and
+      values.\nObservation: the result of the action\n```\n\nOnce all necessary information
+      is gathered, return the following format:\n\n```\nThought: I now know the final
+      answer\nFinal Answer: the final answer to the original input question\n```"},
+      {"role": "user", "content": "\nCurrent Task: Call the iterating tool 5 times\n\nThis
+      is the expected criteria for your final answer: A list of the iterations\nyou
+      MUST return the actual complete content as the final answer, not a summary.\n\nBegin!
+      This is VERY important to you, use the tools available and give your best Final
+      Answer, your job depends on it!\n\nThought:"}, {"role": "assistant", "content":
+      "```\nThought: I need to call the iterating tool the first time.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"First iteration\"}\nObservation: Iteration 0: First
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the first iteration and need to proceed to the second one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Second iteration\"}\nObservation: Iteration 0: Second
+      iteration"}, {"role": "assistant", "content": "```\nThought: I have completed
+      the second iteration and need to proceed to the third one.\nAction: iterating_tool\nAction
+      Input: {\"input_text\": \"Third iteration\"}\nObservation: Iteration 0: Third
+      iteration\n\n\nYou ONLY have access to the following tools, and should NEVER
+      make up tools that are not listed here:\n\nTool Name: iterating_tool\nTool Arguments:
+      {''input_text'': {''description'': None, ''type'': ''str''}}\nTool Description:
+      A tool that iterates a given number of times\n\nIMPORTANT: Use the following
+      format in your response:\n\n```\nThought: you should always think about what
+      to do\nAction: the action to take, only one name of [iterating_tool], just the
+      name, exactly as it''s written.\nAction Input: the input to the action, just
+      a simple JSON object, enclosed in curly braces, using \" to wrap keys and values.\nObservation:
+      the result of the action\n```\n\nOnce all necessary information is gathered,
+      return the following format:\n\n```\nThought: I now know the final answer\nFinal
+      Answer: the final answer to the original input question\n```"}, {"role": "assistant",
+      "content": "```\nThought: I have completed the third iteration and need to proceed
+      to the fourth one.\nAction: iterating_tool\nAction Input: {\"input_text\": \"Fourth
+      iteration\"}\nObservation: Iteration 0: Fourth iteration"}, {"role": "assistant",
+      "content": "```\nThought: I have completed the fourth iteration and need to
+      proceed to the fifth one.\nAction: iterating_tool\nAction Input: {\"input_text\":
+      \"Fifth iteration\"}\nObservation: Iteration 0: Fifth iteration"}], "model":
+      "gpt-4o", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate, zstd
+      connection:
+      - keep-alive
+      content-length:
+      - '3514'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=6OZC5kPO9TEmVSACP2sSOZK9ZEZ5I4T_VUlfmzsoY9Y-1755623255-1.0.1.1-ek3SaNBOhXmCg7K3J7LIsE0aCrnK5YfSumHDT6nc8Df1Zh3bzMLHLDqTUwtqiG8SwxiIFXeGP4.Vt2sx9b3FCkxoyrqNpgrBL5DAffAGHm8;
+        _cfuvid=R_H7SrOF3QFEWePZfvxzuyKWZAt5ulsNbP28.6DC9wM-1755623255760-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.93.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.93.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-read-timeout:
+      - '600.0'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.9
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    body:
+      string: !!binary |
+        H4sIAAAAAAAAA4xTy26jMBTd8xVXXpOK0IRE7KqRInVazaZZTROBY1/AU2NT26QzqvLvI0MSSNuJ
+        ZsPC58E99/EeABDBSQqEVdSxupGTb8lDcne7pt8Xb/r14XX5tJiufiT1z8f944smoVfo3S9k7qS6
+        YbpuJDqhVQ8zg9Shd50u5vMkvo2TuANqzVF6Wdm4yUxP4iieTaLlJEqOwkoLhpak8BwAALx3X1+i
+        4vibpBCFp5caraUlkvRMAiBGS/9CqLXCOqocCQeQaeVQdVXneb5R60q3ZeVSuIeK7hGOKZADlRLm
+        IBwa6kPZm41aCUUl3Cn7hiaF5w25P6EQpbASxrpBsCEhfGA8IdOKX6WsK2GuM1a6Na66ThHFJWO7
+        UXmej/tgsGgt9WNQrZQjgCqlXZ/YT2B7RA7nnktdNkbv7AcpKYQStsoMUquV7691uiEdeggAtt1s
+        24txkcbounGZ0y/Y/W4Rz3s/MmzTgM6TI+i0o3KkWk7DL/wyjo4KaUfbQRhlFfJBOqwSbbnQIyAY
+        pf5czVfefXKhyv+xHwDGsHHIs8YgF+wy8UAz6I/tX7Rzl7uCiUWzFwwzJ9D4SXAsaCv7OyD2j3VY
+        Z4VQJZrGiP4YiiabJbuiiDBiSxIcgr8AAAD//wMAnPTcwxUEAAA=
+    headers:
+      CF-RAY:
+      - 971b3faa6ba46897-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 19 Aug 2025 17:07:43 GMT
+      Server:
+      - cloudflare
+      Strict-Transport-Security:
+      - max-age=31536000; includeSubDomains; preload
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1392'
+      openai-project:
+      - proj_xitITlrFeen7zjNSzML82h9x
+      openai-version:
+      - '2020-10-01'
+      x-envoy-upstream-service-time:
+      - '1841'
+      x-ratelimit-limit-requests:
+      - '10000'
+      x-ratelimit-limit-tokens:
+      - '30000000'
+      x-ratelimit-remaining-requests:
+      - '9999'
+      x-ratelimit-remaining-tokens:
+      - '29999217'
+      x-ratelimit-reset-requests:
+      - 6ms
+      x-ratelimit-reset-tokens:
+      - 1ms
+      x-request-id:
+      - req_19dc255cec9d4763be7d5f597c80e936
+    status:
+      code: 200
+      message: OK
+version: 1

tests/flow_test.py

@@ -497,6 +497,131 @@ def test_unstructured_flow_event_emission():
     assert isinstance(received_events[6].timestamp, datetime)
 
 
+def test_flow_trigger_payload_injection():
+    captured_payload = []
+
+    class TriggerFlow(Flow):
+        @start()
+        def start_method(self, crewai_trigger_payload=None):
+            captured_payload.append(crewai_trigger_payload)
+            return "started"
+
+        @listen(start_method)
+        def second_method(self):
+            captured_payload.append("no_parameter")
+            return "finished"
+
+    flow = TriggerFlow()
+
+    test_payload = "This is important trigger data"
+    flow.kickoff(inputs={"crewai_trigger_payload": test_payload})
+
+    assert captured_payload == [test_payload, "no_parameter"]
+
+
+def test_flow_trigger_payload_injection_multiple_starts():
+    captured_payloads = []
+
+    class MultiStartFlow(Flow):
+        @start()
+        def start_method_1(self, crewai_trigger_payload=None):
+            captured_payloads.append(("start_1", crewai_trigger_payload))
+            return "start_1_done"
+
+        @start()
+        def start_method_2(self, crewai_trigger_payload=None):
+            captured_payloads.append(("start_2", crewai_trigger_payload))
+            return "start_2_done"
+
+    flow = MultiStartFlow()
+
+    test_payload = "Multiple start trigger data"
+    flow.kickoff(inputs={"crewai_trigger_payload": test_payload})
+
+    assert captured_payloads == [("start_1", test_payload), ("start_2", test_payload)]
+
+
+def test_flow_without_trigger_payload():
+    captured_payload = []
+
+    class NormalFlow(Flow):
+        @start()
+        def start_method(self, crewai_trigger_payload=None):
+            captured_payload.append(crewai_trigger_payload)
+            return "no_trigger"
+
+    flow = NormalFlow()
+
+    flow.kickoff(inputs={"other_data": "some value"})
+
+    assert captured_payload[0] is None
+
+
+def test_flow_trigger_payload_with_structured_state():
+    class TriggerState(BaseModel):
+        id: str = "test"
+        message: str = ""
+
+    class StructuredFlow(Flow[TriggerState]):
+        @start()
+        def start_method(self, crewai_trigger_payload=None):
+            return crewai_trigger_payload
+
+    flow = StructuredFlow()
+
+    test_payload = "Structured state trigger data"
+    result = flow.kickoff(inputs={"crewai_trigger_payload": test_payload})
+
+    assert result == test_payload
+
+
+def test_flow_start_method_without_trigger_parameter():
+    execution_order = []
+
+    class FlowWithoutParameter(Flow):
+        @start()
+        def start_without_param(self):
+            execution_order.append("start_executed")
+            return "started"
+
+        @listen(start_without_param)
+        def second_method(self):
+            execution_order.append("second_executed")
+            return "finished"
+
+    flow = FlowWithoutParameter()
+
+    result = flow.kickoff(inputs={"crewai_trigger_payload": "some data"})
+
+    assert execution_order == ["start_executed", "second_executed"]
+    assert result == "finished"
+
+
+def test_async_flow_with_trigger_payload():
+    captured_payload = []
+
+    class AsyncTriggerFlow(Flow):
+        @start()
+        async def async_start_method(self, crewai_trigger_payload=None):
+            captured_payload.append(crewai_trigger_payload)
+            await asyncio.sleep(0.01)
+            return "async_started"
+
+        @listen(async_start_method)
+        async def async_second_method(self, result):
+            captured_payload.append(result)
+            await asyncio.sleep(0.01)
+            return "async_finished"
+
+    flow = AsyncTriggerFlow()
+
+    test_payload = "Async trigger data"
+    result = asyncio.run(flow.kickoff_async(inputs={"crewai_trigger_payload": test_payload}))
+
+    assert captured_payload == [test_payload, "async_started"]
+    assert result == "async_finished"
+
+
 def test_structured_flow_event_emission():
     """Test that the correct events are emitted during structured flow
     execution with all fields validated."""

tests/tools/test_base_tool.py

@@ -1,9 +1,12 @@
 import asyncio
-import inspect
-import unittest
-from typing import Any, Callable, Dict, List
+from typing import Callable
 from unittest.mock import patch
 
+import pytest
+
+from crewai.agent import Agent
+from crewai.crew import Crew
+from crewai.task import Task
 from crewai.tools import BaseTool, tool
 
 
@@ -130,6 +133,7 @@ def test_result_as_answer_in_tool_decorator():
 
 class SyncTool(BaseTool):
     """Test implementation with a synchronous _run method"""
+
     name: str = "sync_tool"
     description: str = "A synchronous tool for testing"
 
@@ -140,6 +144,7 @@ class SyncTool(BaseTool):
 
 class AsyncTool(BaseTool):
     """Test implementation with an asynchronous _run method"""
+
     name: str = "async_tool"
     description: str = "An asynchronous tool for testing"
 
@@ -174,7 +179,7 @@ def test_run_calls_asyncio_run_for_async_tools():
     """Test that asyncio.run is called when using async tools."""
     async_tool = AsyncTool()
 
-    with patch('asyncio.run') as mock_run:
+    with patch("asyncio.run") as mock_run:
         mock_run.return_value = "Processed test asynchronously"
         async_result = async_tool.run(input_text="test")
 
@@ -186,9 +191,43 @@ def test_run_does_not_call_asyncio_run_for_sync_tools():
     """Test that asyncio.run is NOT called when using sync tools."""
     sync_tool = SyncTool()
 
-    with patch('asyncio.run') as mock_run:
+    with patch("asyncio.run") as mock_run:
         sync_result = sync_tool.run(input_text="test")
 
         mock_run.assert_not_called()
         assert sync_result == "Processed test synchronously"
 
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_max_usage_count_is_respected():
+    class IteratingTool(BaseTool):
+        name: str = "iterating_tool"
+        description: str = "A tool that iterates a given number of times"
+
+        def _run(self, input_text: str):
+            return f"Iteration {input_text}"
+
+    tool = IteratingTool(max_usage_count=5)
+
+    agent = Agent(
+        role="Iterating Agent",
+        goal="Call the iterating tool 5 times",
+        backstory="You are an agent that iterates a given number of times",
+        tools=[tool],
+    )
+
+    task = Task(
+        description="Call the iterating tool 5 times",
+        expected_output="A list of the iterations",
+        agent=agent,
+    )
+
+    crew = Crew(
+        agents=[agent],
+        tasks=[task],
+        verbose=True,
+    )
+
+    crew.kickoff()
+    assert tool.max_usage_count == 5
+    assert tool.current_usage_count == 5

uv.lock

@@ -1,5 +1,5 @@
 version = 1
-revision = 2
+revision = 3
 requires-python = ">=3.10, <3.14"
 resolution-markers = [
     "python_full_version >= '3.13' and platform_python_implementation == 'PyPy' and sys_platform == 'darwin'",
@@ -787,7 +787,7 @@ requires-dist = [
     { name = "litellm", specifier = "==1.74.9" },
     { name = "mem0ai", marker = "extra == 'mem0'", specifier = ">=0.1.94" },
     { name = "onnxruntime", specifier = "==1.22.0" },
-    { name = "openai", specifier = "<1.100.0" },
+    { name = "openai", specifier = ">=1.13.3" },
     { name = "openpyxl", specifier = ">=3.1.5" },
     { name = "openpyxl", marker = "extra == 'openpyxl'", specifier = ">=3.1.5" },
     { name = "opentelemetry-api", specifier = ">=1.30.0" },