vectara-agentic 0.2.13__tar.gz → 0.2.15__tar.gz

{vectara_agentic-0.2.13/vectara_agentic.egg-info → vectara_agentic-0.2.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vectara_agentic
-Version: 0.2.13
+Version: 0.2.15
 Summary: A Python package for creating AI Assistants and AI Agents with Vectara
 Home-page: https://github.com/vectara/py-vectara-agentic
 Author: Ofer Mendelevitch
@@ -16,18 +16,18 @@ Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: llama-index==0.12.31
-Requires-Dist: llama-index-indices-managed-vectara==0.4.3
+Requires-Dist: llama-index==0.12.33
+Requires-Dist: llama-index-indices-managed-vectara==0.4.4
 Requires-Dist: llama-index-agent-llm-compiler==0.3.0
 Requires-Dist: llama-index-agent-lats==0.3.0
 Requires-Dist: llama-index-agent-openai==0.4.6
-Requires-Dist: llama-index-llms-openai==0.3.35
+Requires-Dist: llama-index-llms-openai==0.3.38
 Requires-Dist: llama-index-llms-anthropic==0.6.10
 Requires-Dist: llama-index-llms-together==0.3.1
 Requires-Dist: llama-index-llms-groq==0.3.1
 Requires-Dist: llama-index-llms-fireworks==0.3.2
 Requires-Dist: llama-index-llms-cohere==0.4.1
-Requires-Dist: llama-index-llms-gemini==0.4.14
+Requires-Dist: llama-index-llms-google-genai==0.1.8
 Requires-Dist: llama-index-llms-bedrock==0.3.8
 Requires-Dist: llama-index-tools-yahoo-finance==0.3.0
 Requires-Dist: llama-index-tools-arxiv==0.3.0
@@ -49,7 +49,7 @@ Requires-Dist: arize-phoenix==8.26.1
 Requires-Dist: arize-phoenix-otel==0.9.2
 Requires-Dist: protobuf==5.29.3
 Requires-Dist: tokenizers>=0.20
-Requires-Dist: pydantic==2.10.6
+Requires-Dist: pydantic==2.11.3
 Requires-Dist: retrying==1.3.4
 Requires-Dist: python-dotenv==1.0.1
 Requires-Dist: tiktoken==0.9.0
@@ -292,13 +292,12 @@ For example, in the quickstart example the schema is:
 
 ```python
 class QueryFinancialReportsArgs(BaseModel):
-    query: str = Field(..., description="The user query.")
     year: int | str = Field(..., description=f"The year this query relates to. An integer between {min(years)} and {max(years)} or a string specifying a condition on the year (example: '>2020').")
     ticker: str = Field(..., description=f"The company ticker. Must be a valid ticket symbol from the list {tickers.keys()}.")
 ```
 
-The `query` is required and is always the query string.
-The other arguments are optional and will be interpreted as Vectara metadata filters.
+Remember, the `query` argument is part of the rag_tool that is generated, but `vectara-agentic` creates it and you do 
+not need to specify it explicitly. 
 
 For example, in the example above, the agent may call the `query_financial_reports_tool` tool with 
 query='what is the revenue?', year=2022 and ticker='AAPL'. Subsequently the RAG tool will issue
@@ -483,13 +482,28 @@ class MyWorkflow(Workflow):
         return StopEvent(result="Hello, world!")
 ```
 
-When the `run()` method in vectara-agentic is invoked, it calls the workflow with the following variables in the StartEvent:
+When the `run()` method in vectara-agentic is invoked, it calls the workflow with the following variables in the `StartEvent`:
 * `agent`: the agent object used to call `run()` (self)
 * `tools`: the tools provided to the agent. Those can be used as needed in the flow.
 * `llm`: a pointer to a LlamaIndex llm, so it can be used in the workflow. For example, one of the steps may call `llm.acomplete(prompt)`
 * `verbose`: controls whether extra debug information is displayed
 * `inputs`: this is the actual inputs to the workflow provided by the call to `run()` and must be of type `InputsModel`
 
+If you want to use `agent`, `tools`, `llm` or `verbose` in other events (that are not `StartEvent`), you can store them in
+the `Context` of the Workflow as follows:
+
+```python
+await ctx.set("agent", ev.agent)
+```
+
+and then in any other event you can pull that agent object with
+
+```python
+agent = await ctx.get("agent")
+```
+
+Similarly you can reuse the `llm`, `tools` or `verbose` arguments within other nodes in the workflow.
+
 ### Using the Workflow with Your Agent
 
 When initializing your agent, pass the workflow class using the `workflow_cls` parameter:
@@ -521,7 +535,7 @@ print(workflow_result.answer)
 
 ### Built-in Workflows
 
-`vectara-agentic` includes two powerful workflow implementations that you can use right away:
+`vectara-agentic` includes two workflow implementations that you can use right away:
 
 #### 1. `SubQuestionQueryWorkflow`
 

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/README.md

@@ -222,13 +222,12 @@ For example, in the quickstart example the schema is:
 
 ```python
 class QueryFinancialReportsArgs(BaseModel):
-    query: str = Field(..., description="The user query.")
     year: int | str = Field(..., description=f"The year this query relates to. An integer between {min(years)} and {max(years)} or a string specifying a condition on the year (example: '>2020').")
     ticker: str = Field(..., description=f"The company ticker. Must be a valid ticket symbol from the list {tickers.keys()}.")
 ```
 
-The `query` is required and is always the query string.
-The other arguments are optional and will be interpreted as Vectara metadata filters.
+Remember, the `query` argument is part of the rag_tool that is generated, but `vectara-agentic` creates it and you do 
+not need to specify it explicitly. 
 
 For example, in the example above, the agent may call the `query_financial_reports_tool` tool with 
 query='what is the revenue?', year=2022 and ticker='AAPL'. Subsequently the RAG tool will issue
@@ -413,13 +412,28 @@ class MyWorkflow(Workflow):
         return StopEvent(result="Hello, world!")
 ```
 
-When the `run()` method in vectara-agentic is invoked, it calls the workflow with the following variables in the StartEvent:
+When the `run()` method in vectara-agentic is invoked, it calls the workflow with the following variables in the `StartEvent`:
 * `agent`: the agent object used to call `run()` (self)
 * `tools`: the tools provided to the agent. Those can be used as needed in the flow.
 * `llm`: a pointer to a LlamaIndex llm, so it can be used in the workflow. For example, one of the steps may call `llm.acomplete(prompt)`
 * `verbose`: controls whether extra debug information is displayed
 * `inputs`: this is the actual inputs to the workflow provided by the call to `run()` and must be of type `InputsModel`
 
+If you want to use `agent`, `tools`, `llm` or `verbose` in other events (that are not `StartEvent`), you can store them in
+the `Context` of the Workflow as follows:
+
+```python
+await ctx.set("agent", ev.agent)
+```
+
+and then in any other event you can pull that agent object with
+
+```python
+agent = await ctx.get("agent")
+```
+
+Similarly you can reuse the `llm`, `tools` or `verbose` arguments within other nodes in the workflow.
+
 ### Using the Workflow with Your Agent
 
 When initializing your agent, pass the workflow class using the `workflow_cls` parameter:
@@ -451,7 +465,7 @@ print(workflow_result.answer)
 
 ### Built-in Workflows
 
-`vectara-agentic` includes two powerful workflow implementations that you can use right away:
+`vectara-agentic` includes two workflow implementations that you can use right away:
 
 #### 1. `SubQuestionQueryWorkflow`
 

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/requirements.txt

@@ -1,15 +1,15 @@
-llama-index==0.12.31
-llama-index-indices-managed-vectara==0.4.3
+llama-index==0.12.33
+llama-index-indices-managed-vectara==0.4.4
 llama-index-agent-llm-compiler==0.3.0
 llama-index-agent-lats==0.3.0
 llama-index-agent-openai==0.4.6
-llama-index-llms-openai==0.3.35
+llama-index-llms-openai==0.3.38
 llama-index-llms-anthropic==0.6.10
 llama-index-llms-together==0.3.1
 llama-index-llms-groq==0.3.1
 llama-index-llms-fireworks==0.3.2
 llama-index-llms-cohere==0.4.1
-llama-index-llms-gemini==0.4.14
+llama-index-llms-google-genai ==0.1.8
 llama-index-llms-bedrock==0.3.8
 llama-index-tools-yahoo-finance==0.3.0
 llama-index-tools-arxiv==0.3.0
@@ -31,7 +31,7 @@ arize-phoenix==8.26.1
 arize-phoenix-otel==0.9.2
 protobuf==5.29.3
 tokenizers>=0.20
-pydantic==2.10.6
+pydantic==2.11.3
 retrying==1.3.4
 python-dotenv==1.0.1
 tiktoken==0.9.0

vectara_agentic-0.2.15/tests/test_groq.py

@@ -0,0 +1,120 @@
+import unittest
+
+from pydantic import Field, BaseModel
+
+from vectara_agentic.agent import Agent, AgentType
+from vectara_agentic.agent_config import AgentConfig
+from vectara_agentic.tools import VectaraToolFactory
+from vectara_agentic.types import ModelProvider
+
+
+import nest_asyncio
+nest_asyncio.apply()
+
+tickers = {
+    "C": "Citigroup",
+    "COF": "Capital One",
+    "JPM": "JPMorgan Chase",
+    "AAPL": "Apple Computer",
+    "GOOG": "Google",
+    "AMZN": "Amazon",
+    "SNOW": "Snowflake",
+    "TEAM": "Atlassian",
+    "TSLA": "Tesla",
+    "NVDA": "Nvidia",
+    "MSFT": "Microsoft",
+    "AMD": "Advanced Micro Devices",
+    "INTC": "Intel",
+    "NFLX": "Netflix",
+    "STT": "State Street",
+    "BK": "Bank of New York Mellon",
+}
+years = list(range(2015, 2025))
+
+
+def mult(x: float, y: float) -> float:
+    "Multiply two numbers"
+    return x * y
+
+
+def get_company_info() -> list[str]:
+    """
+    Returns a dictionary of companies you can query about. Always check this before using any other tool.
+    The output is a dictionary of valid ticker symbols mapped to company names.
+    You can use this to identify the companies you can query about, and their ticker information.
+    """
+    return tickers
+
+
+def get_valid_years() -> list[str]:
+    """
+    Returns a list of the years for which financial reports are available.
+    Always check this before using any other tool.
+    """
+    return years
+
+
+config_gemini = AgentConfig(
+    agent_type=AgentType.FUNCTION_CALLING,
+    main_llm_provider=ModelProvider.GEMINI,
+    tool_llm_provider=ModelProvider.GEMINI,
+)
+
+
+fc_config_groq = AgentConfig(
+    agent_type=AgentType.FUNCTION_CALLING,
+    main_llm_provider=ModelProvider.GROQ,
+    tool_llm_provider=ModelProvider.GROQ,
+)
+
+
+class TestGROQ(unittest.TestCase):
+
+    def test_tool_with_many_arguments(self):
+
+        vectara_corpus_key = "vectara-docs_1"
+        vectara_api_key = "zqt_UXrBcnI2UXINZkrv4g1tQPhzj02vfdtqYJIDiA"
+        vec_factory = VectaraToolFactory(vectara_corpus_key, vectara_api_key)
+
+        class QueryToolArgs(BaseModel):
+            arg1: str = Field(description="the first argument", examples=["val1"])
+            arg2: str = Field(description="the second argument", examples=["val2"])
+            arg3: str = Field(description="the third argument", examples=["val3"])
+            arg4: str = Field(description="the fourth argument", examples=["val4"])
+            arg5: str = Field(description="the fifth argument", examples=["val5"])
+            arg6: str = Field(description="the sixth argument", examples=["val6"])
+            arg7: str = Field(description="the seventh argument", examples=["val7"])
+            arg8: str = Field(description="the eighth argument", examples=["val8"])
+            arg9: str = Field(description="the ninth argument", examples=["val9"])
+            arg10: str = Field(description="the tenth argument", examples=["val10"])
+            arg11: str = Field(description="the eleventh argument", examples=["val11"])
+            arg12: str = Field(description="the twelfth argument", examples=["val12"])
+            arg13: str = Field(
+                description="the thirteenth argument", examples=["val13"]
+            )
+            arg14: str = Field(
+                description="the fourteenth argument", examples=["val14"]
+            )
+            arg15: str = Field(description="the fifteenth argument", examples=["val15"])
+
+        query_tool_1 = vec_factory.create_rag_tool(
+            tool_name="rag_tool",
+            tool_description="""
+            A dummy tool that takes 15 arguments and returns a response (str) to the user query based on the data in this corpus.
+            We are using this tool to test the tool factory works and does not crash with OpenAI.
+            """,
+            tool_args_schema=QueryToolArgs,
+        )
+
+        agent = Agent(
+            tools=[query_tool_1],
+            topic="Sample topic",
+            custom_instructions="Call the tool with 15 arguments",
+            agent_config=fc_config_groq,
+        )
+        res = agent.chat("What is the stock price?")
+        self.assertIn("I don't know", str(res))
+
+
+if __name__ == "__main__":
+    unittest.main()

vectara_agentic-0.2.15/tests/test_return_direct.py

@@ -0,0 +1,49 @@
+import unittest
+
+from vectara_agentic.agent import Agent
+from vectara_agentic.tools import VectaraToolFactory
+
+vectara_corpus_key = "vectara-docs_1"
+vectara_api_key = "zqt_UXrBcnI2UXINZkrv4g1tQPhzj02vfdtqYJIDiA"
+
+
+class TestAgentPackage(unittest.TestCase):
+
+    def test_return_direct1(self):
+        vec_factory = VectaraToolFactory(vectara_corpus_key, vectara_api_key)
+
+        query_tool = vec_factory.create_rag_tool(
+            tool_name="rag_tool",
+            tool_description="""
+            A dummy tool for testing return_direct.
+            """,
+            return_direct=True,
+        )
+
+        agent = Agent(
+            tools=[query_tool],
+            topic="Sample topic",
+            custom_instructions="You are a helpful assistant.",
+        )
+        res = agent.chat("What is Vectara?")
+        self.assertIn("Response:", str(res))
+        self.assertIn("fcs_score", str(res))
+        self.assertIn("References:", str(res))
+
+    def test_from_corpus(self):
+        agent = Agent.from_corpus(
+            tool_name="rag_tool",
+            vectara_corpus_key=vectara_corpus_key,
+            vectara_api_key=vectara_api_key,
+            data_description="stuff about Vectara",
+            assistant_specialty="question answering",
+            return_direct=True,
+        )
+        res = agent.chat("What is Vectara?")
+        self.assertIn("Response:", str(res))
+        self.assertIn("fcs_score", str(res))
+        self.assertIn("References:", str(res))
+
+
+if __name__ == "__main__":
+    unittest.main()

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/tests/test_tools.py

@@ -9,6 +9,7 @@ from vectara_agentic.tools import (
 )
 from vectara_agentic.agent import Agent
 from vectara_agentic.agent_config import AgentConfig
+from vectara_agentic.types import AgentType, ModelProvider
 
 from llama_index.core.tools import FunctionTool
 
@@ -179,22 +180,57 @@ class TestToolsPackage(unittest.TestCase):
         query_tool_1 = vec_factory.create_rag_tool(
             tool_name="rag_tool",
             tool_description="""
-            A dummy tool that takes 20 arguments and returns a response (str) to the user query based on the data in this corpus.
+            A dummy tool that takes 15 arguments and returns a response (str) to the user query based on the data in this corpus.
             We are using this tool to test the tool factory works and does not crash with OpenAI.
             """,
             tool_args_schema=QueryToolArgs,
         )
 
-        config = AgentConfig()
+        # Test with 15 arguments which go over the 1024 limit.
+        config = AgentConfig(
+            agent_type=AgentType.OPENAI
+        )
         agent = Agent(
             tools=[query_tool_1],
             topic="Sample topic",
-            custom_instructions="Call the tool with 20 arguments",
+            custom_instructions="Call the tool with 15 arguments for OPENAI",
             agent_config=config,
         )
-        res = agent.chat("What is the stock price?")
+        res = agent.chat("What is the stock price for Yahoo on 12/31/22?")
         self.assertIn("maximum length of 1024 characters", str(res))
 
+        # Same test but with GROQ
+        config = AgentConfig(
+            agent_type=AgentType.FUNCTION_CALLING,
+            main_llm_provider=ModelProvider.GROQ,
+            tool_llm_provider=ModelProvider.GROQ,
+        )
+        agent = Agent(
+            tools=[query_tool_1],
+            topic="Sample topic",
+            custom_instructions="Call the tool with 15 arguments for GROQ",
+            agent_config=config,
+        )
+        res = agent.chat("What is the stock price?")
+        self.assertNotIn("maximum length of 1024 characters", str(res))
+
+        # Same test but with ANTHROPIC
+        config = AgentConfig(
+            agent_type=AgentType.FUNCTION_CALLING,
+            main_llm_provider=ModelProvider.ANTHROPIC,
+            tool_llm_provider=ModelProvider.ANTHROPIC,
+        )
+        agent = Agent(
+            tools=[query_tool_1],
+            topic="Sample topic",
+            custom_instructions="Call the tool with 15 arguments for ANTHROPIC",
+            agent_config=config,
+        )
+        res = agent.chat("What is the stock price?")
+        # ANTHROPIC does not have that 1024 limit
+        self.assertIn("stock price", str(res))
+
+        # But using Compact_docstring=True, we can pass 15 arguments successfully.
         vec_factory = VectaraToolFactory(
             vectara_corpus_key, vectara_api_key, compact_docstring=True
         )
@@ -211,7 +247,7 @@ class TestToolsPackage(unittest.TestCase):
         agent = Agent(
             tools=[query_tool_2],
             topic="Sample topic",
-            custom_instructions="Call the tool with 20 arguments",
+            custom_instructions="Call the tool with 15 arguments",
             agent_config=config,
         )
         res = agent.chat("What is the stock price?")
@@ -227,7 +263,7 @@ class TestToolsPackage(unittest.TestCase):
             tool_name="ask_vectara",
             data_description="data from Vectara website",
             assistant_specialty="RAG as a service",
-            vectara_summarizer="mockingbird-1.0-2024-07-16",
+            vectara_summarizer="mockingbird-2.0",
         )
 
         self.assertIn(

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/tests/test_vectara_llms.py

@@ -15,7 +15,10 @@ vectara_api_key = "zqt_UXrBcnI2UXINZkrv4g1tQPhzj02vfdtqYJIDiA"
 class TestLLMPackage(unittest.TestCase):
 
     def test_vectara_openai(self):
-        vec_factory = VectaraToolFactory(vectara_corpus_key, vectara_api_key)
+        vec_factory = VectaraToolFactory(
+            vectara_corpus_key=vectara_corpus_key,
+            vectara_api_key=vectara_api_key
+        )
 
         self.assertEqual(vectara_corpus_key, vec_factory.vectara_corpus_key)
         self.assertEqual(vectara_api_key, vec_factory.vectara_api_key)
@@ -51,17 +54,6 @@ class TestLLMPackage(unittest.TestCase):
 
     def test_vectara_mockingbird(self):
         vec_factory = VectaraToolFactory(vectara_corpus_key, vectara_api_key)
-
-        query_tool = vec_factory.create_rag_tool(
-            tool_name="rag_tool",
-            tool_description="""
-            Returns a response (str) to the user query based on the data in this corpus.
-            """,
-            vectara_summarizer="mockingbird-1.0-2024-07-16",
-        )
-        res = query_tool(query="What is Vectara?")
-        self.assertIn("Vectara is an end-to-end platform", str(res))
-
         query_tool = vec_factory.create_rag_tool(
             tool_name="rag_tool",
             tool_description="""

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/vectara_agentic/_observability.py

@@ -12,28 +12,50 @@ def setup_observer(config: AgentConfig, verbose: bool) -> bool:
     '''
     Setup the observer.
     '''
-    import phoenix as px
-    from openinference.instrumentation.llama_index import LlamaIndexInstrumentor
-    from phoenix.otel import register
-    if config.observer == ObserverType.ARIZE_PHOENIX:
-        phoenix_endpoint = os.getenv("PHOENIX_ENDPOINT", None)
-        if not phoenix_endpoint:
-            px.launch_app()
-            tracer_provider = register(endpoint='http://localhost:6006/v1/traces', project_name="vectara-agentic")
-        elif 'app.phoenix.arize.com' in phoenix_endpoint:   # hosted on Arize
-            phoenix_api_key = os.getenv("PHOENIX_API_KEY", None)
-            if not phoenix_api_key:
-                raise ValueError("Arize Phoenix API key not set. Please set PHOENIX_API_KEY environment variable.")
-            os.environ["PHOENIX_CLIENT_HEADERS"] = f"api_key={phoenix_api_key}"
-            os.environ["PHOENIX_COLLECTOR_ENDPOINT"] = "https://app.phoenix.arize.com"
-            tracer_provider = register(endpoint=phoenix_endpoint, project_name="vectara-agentic")
-        else:       # Self hosted Phoenix
-            tracer_provider = register(endpoint=phoenix_endpoint, project_name="vectara-agentic")
-        LlamaIndexInstrumentor().instrument(tracer_provider=tracer_provider)
-        return True
+    if config.observer != ObserverType.ARIZE_PHOENIX:
+        if verbose:
+            print("No Phoenix observer set.")
+        return False
+
+    try:
+        import phoenix as px
+        from openinference.instrumentation.llama_index import LlamaIndexInstrumentor
+        from phoenix.otel import register
+    except ImportError:
+        print(
+            (
+                "Phoenix libraries not found. Please install with"
+                "'pip install arize-phoenix openinference-instrumentation-llama-index'"
+            )
+        )
+        return False
+
+    phoenix_endpoint = os.getenv("PHOENIX_ENDPOINT", None)
+    if not phoenix_endpoint:
+        print("Phoenix endpoint not set. Attempting to launch local Phoenix UI...")
+        px.launch_app()
+        print("Local Phoenix UI launched. You can view traces at the UI address (usually http://localhost:6006).")
+
+    if phoenix_endpoint and 'app.phoenix.arize.com' in phoenix_endpoint:
+        phoenix_api_key = os.getenv("PHOENIX_API_KEY")
+        if not phoenix_api_key:
+            raise ValueError(
+                "Arize Phoenix API key not set. Please set PHOENIX_API_KEY."
+            )
+        os.environ["PHOENIX_CLIENT_HEADERS"] = f"api_key={phoenix_api_key}"
+        os.environ["PHOENIX_COLLECTOR_ENDPOINT"] = "https://app.phoenix.arize.com"
+
+    reg_kwargs = {
+        "endpoint": phoenix_endpoint or 'http://localhost:6006/v1/traces',
+        "project_name": "vectara-agentic",
+        "batch": True,
+        "set_global_tracer_provider": False,
+    }
+    tracer_provider = register(**reg_kwargs)
+    LlamaIndexInstrumentor().instrument(tracer_provider=tracer_provider)
     if verbose:
-        print("No observer set.")
-    return False
+        print(f"Phoenix observer configured for project 'vectara-agentic' at endpoint: {reg_kwargs['endpoint']}")
+    return True
 
 
 def _extract_fcs_value(output: Union[str, dict]) -> Optional[float]:

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/vectara_agentic/_prompts.py

@@ -5,7 +5,7 @@ This file contains the prompt templates for the different types of agents.
 # General (shared) instructions
 GENERAL_INSTRUCTIONS = """
 - Use tools as your main source of information, do not respond without using a tool. Do not respond based on pre-trained knowledge.
-- Use the 'get_bad_topics' tool to determine the topics you are not allowed to discuss or respond to.
+- Use the 'get_bad_topics' (if it exists) tool to determine the topics you are not allowed to discuss or respond to.
 - Before responding to a user query that requires knowledge of the current date, call the 'get_current_date' tool to get the current date.
   Never rely on previous knowledge of the current date.
   Example queries that require the current date: "What is the revenue of Apple last october?" or "What was the stock price 5 days ago?".

{vectara_agentic-0.2.13 → vectara_agentic-0.2.15}/vectara_agentic/_version.py

@@ -1,4 +1,4 @@
 """
 Define the version of the package.
 """
-__version__ = "0.2.13"
+__version__ = "0.2.15"