fast-agent-mcp 0.2.36__py3-none-any.whl → 0.2.37__py3-none-any.whl

mcp_agent/resources/examples/in_dev/slides.py

@@ -1,110 +0,0 @@
-import asyncio
-from pathlib import Path
-from typing import TYPE_CHECKING
-
-from mcp_agent.core.fastagent import FastAgent
-from mcp_agent.llm.augmented_llm import RequestParams
-from mcp_agent.mcp.prompts.prompt_load import load_prompt_multipart
-
-if TYPE_CHECKING:
-    from mcp_agent.mcp.prompt_message_multipart import PromptMessageMultipart
-
-# Create the application
-fast = FastAgent("Data Analysis (Roots)")
-
-
-# The sample data is under Database Contents License (DbCL) v1.0.
-
-# Available here : https://www.kaggle.com/datasets/pavansubhasht/ibm-hr-analytics-attrition-dataset
-
-# The CSS files are distributed under the MIT License from the excellent
-# marpstyle project : https://github.com/cunhapaulo/marpstyle
-
-
-@fast.agent(
-    name="slides",
-    servers=["filesystem"],
-    instruction="""
-You produce compelling slide decks for impactful presentations. You usually try and keep the pack to between
-8-12 slides, with key insights at the start, backed up with data, diagrams and analysis as available. You 
-are able to help direct colour, style and and questions for enhancing the presentation. Produced charts and
-visualisations will be in the ./mount-point/ directory. You output MARP markdown files.
-""",
-    request_params=RequestParams(maxTokens=8192),
-)
-@fast.agent(
-    name="data_analysis",
-    instruction="""
-You have access to a Python 3.12 interpreter and you can use this to analyse and process data. 
-Common analysis packages such as Pandas, Seaborn and Matplotlib are already installed. 
-You can add further packages if needed.
-Data files are accessible from the /mnt/data/ directory (this is the current working directory).
-Visualisations should be saved as .png files in the current working directory.
-""",
-    servers=["interpreter"],
-    request_params=RequestParams(maxTokens=8192),
-)
-@fast.orchestrator(
-    name="orchestrator",
-    plan_type="iterative",
-    agents=["slides", "data_analysis"],
-)
-async def main() -> None:
-    # Use the app's context manager
-    async with fast.run() as agent:
-        prompts: list[PromptMessageMultipart] = load_prompt_multipart(Path("slides.md"))
-        await agent.slides.apply_prompt_messages(prompts)
-
-        await agent.orchestrator.send(
-            "Produce a compelling presentation for the CSV data file in the /mnt/data/ directory."
-            "The slides agent will produce a presentation, make sure to consult it first for "
-            "colour scheme and formatting guidance. Make sure that any 'call-outs' have a distinct"
-            "background to ensure they stand out."
-            "Make sure the presentation is impactful, concise and visualises key insights in to the data"
-            " in a compelling way."
-            "The presentation is by the 'llmindset team' and produced in 'march 2025'."
-            "Produce it step-by-step; long responses without checking in are likely to exceed"
-            "maximum output token limits."
-        )
-        # colours: str = await agent.slides.send("Tell the Data Analysis agent what colour schemes and chart sizes you prefer for the presentation")
-
-        # analysis: str = await agent.data_analysis.send(
-        #     "Examine the CSV file in /mnt/data, produce a detailed analysis of the data,"
-        #     "and any patterns it contains. Visualise some of the key points, saving .png files to"
-        #     "your current workig folder (/mnt/data). Respond with a summary of your findings, and a list"
-        #     "of visualiations and their filenames ready to incorporate in to a slide deck. The presentation agent has"
-        #     f"specified the following style guidelines for generated charts:\n {colours}"
-        # )
-        # await agent.slides.send(
-        #     "Produce a MARP Presentation for the this analysis. You will find the visualisations in "
-        #     f"in the ./mount-point/ folder. The analysis follows....\n{analysis}"
-        # )
-
-        await agent()
-
-
-if __name__ == "__main__":
-    asyncio.run(main())
-
-
-############################################################################################################
-# Example of evaluator/optimizer flow
-############################################################################################################
-# @fast.agent(
-#     "evaluator",
-#     """You are collaborating with a Data Analysis tool that has the capability to analyse data and produce visualisations.
-#     You must make sure that the tool has:
-#      - Considered the best way for a Human to interpret the data
-#      - Produced insightful visualasions.
-#      - Provided a high level summary report for the Human.
-#      - Has had its findings challenged, and justified
-#     """,
-#     request_params=RequestParams(maxTokens=8192),
-# )
-# @fast.evaluator_optimizer(
-#     "analysis_tool",
-#     generator="data_analysis",
-#     evaluator="evaluator",
-#     max_refinements=3,
-#     min_rating="EXCELLENT",
-# )

mcp_agent/resources/examples/internal/agent.py

@@ -1,20 +0,0 @@
-import asyncio
-
-from mcp_agent.core.fastagent import FastAgent
-
-# Create the application
-fast = FastAgent("FastAgent Example")
-
-
-# Define the agent
-@fast.agent(servers=["category", "mcp_hfspace","mcp_webcam"])
-#@fast.agent(name="test")
-async def main() -> None:
-    # use the --model command line switch or agent arguments to change model
-    async with fast.run() as agent:
-#        await agent.prompt(agent_name="test")
-        await agent()
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

mcp_agent/resources/examples/internal/fastagent.config.yaml

@@ -1,66 +0,0 @@
-default_model: sonnet
-
-# on windows, adjust the mount point to be the full path e.g. x:/temp/data-analysis/mount-point:/mnt/data/
-
-# logger:
-#   level: "debug"
-#   type: "console"
-
-mcp:
-  servers:
-    interpreter:
-      command: "docker"
-      args:
-        [
-          "run",
-          "-i",
-          "--rm",
-          "--pull=always",
-          "-v",
-          "./mount-point:/mnt/data/",
-          "ghcr.io/evalstate/mcp-py-repl:latest",
-        ]
-      roots:
-        - uri: "file://./mount-point/"
-          name: "test_data"
-          server_uri_alias: "file:///mnt/data/"
-    filesystem:
-      # On windows update the command and arguments to use `node` and the absolute path to the server.
-      # Use `npm i -g @modelcontextprotocol/server-filesystem` to install the server globally.
-      # Use `npm -g root` to find the global node_modules path.`
-      # command: "node"
-      # args: ["c:/Program Files/nodejs/node_modules/@modelcontextprotocol/server-filesystem/dist/index.js","."]
-      command: "npx"
-      args:
-        [
-          "-y",
-          "@modelcontextprotocol/server-filesystem",
-          "src/mcp_agent/resources/examples/data-analysis/mount-point/",
-        ]
-    fetch:
-      command: "uvx"
-      args: ["mcp-server-fetch"]
-    brave:
-      # On windows replace the command and args line to use `node` and the absolute path to the server.
-      # Use `npm i -g @modelcontextprotocol/server-brave-search` to install the server globally.
-      # Use `npm -g root` to find the global node_modules path.`
-      # command: "node"
-      # args: ["c:/Program Files/nodejs/node_modules/@modelcontextprotocol/server-brave-search/dist/index.js"]
-      command: "npx"
-      args: ["-y", "@modelcontextprotocol/server-brave-search"]
-    sizing_setup:
-      command: "uv"
-      args: ["run", "prompt_sizing1.py"]
-
-    category:
-      command: "prompt-server"
-      args: ["simple.txt"]
-
-    mcp_hfspace:
-      command: "npx"
-      args: ["@llmindset/mcp-hfspace"]
-
-    mcp_webcam:
-      command: "npx"
-      args: ["@llmindset/mcp-webcam"]
-

mcp_agent/resources/examples/internal/history_transfer.py

@@ -1,35 +0,0 @@
-import asyncio
-
-from mcp_agent.core.fastagent import FastAgent
-
-# Create the application
-fast = FastAgent("FastAgent Example")
-
-
-# Define the agent
-@fast.agent(name="haiku", model="haiku")
-@fast.agent(name="openai", model="o3-mini.medium")
-
-# @fast.agent(name="test")
-async def main() -> None:
-    async with fast.run() as agent:
-        # Start an interactive session with "haiku"
-        await agent.prompt(agent_name="haiku")
-        # Transfer the message history top "openai"
-        await agent.openai.generate(agent.haiku.message_history)
-        # Continue the conversation
-        await agent.prompt(agent_name="openai")  # Interactive shell
-
-        # result: str = await agent.send("foo")
-        # mcp_prompt: PromptMessage = PromptMessage(
-        #     role="user", content=TextContent(type="text", text="How are you?")
-        # )
-        # result: str = agent.send(mcp_prompt)
-        # resource: ReadResourceResult = agent.openai.get_resource(
-        #     "server_name", "resource://images/cat.png"
-        # )
-        # response: str = Prompt.user("What is in this image?", resource)
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

mcp_agent/resources/examples/internal/job.py

@@ -1,84 +0,0 @@
-"""
-PMO Job Description Generator Agent
-  Purpose: Generate comprehensive PMO job descriptions using a multi-stage approach
-  for clarity, consistency and quality control
-"""
-
-import asyncio
-
-from mcp_agent.core.fastagent import FastAgent
-
-# Create the application
-fast = FastAgent("PMO Job Description Generator")
-
-
-@fast.agent(
-    name="content_generator",
-    instruction="""You are a PMO job description expert. Generate job descriptions for PMO roles         
-    following these guidelines:                                                                          
-    - Focus on modern lean/agile and product-based approaches                                            
-    - Emphasize practical experience and demonstrated results over formal requirements                   
-    - Ensure clear role differentiation with minimal overlap                                             
-    - Format output in Markdown                                                                          
-    - Context: Telecommunications industry in open organization valuing practical experience             
-                                                                                                        
-    Structure each job description with:                                                                 
-    1. Role Title                                                                                        
-    2. Position Summary                                                                                  
-    3. Key Responsibilities                                                                              
-    4. Required Experience                                                                               
-    5. Desired Capabilities                                                                              
-    """,
-    model="anthropic.claude-3-5-haiku-latest",
-)
-@fast.agent(
-    name="consistency_checker",
-    instruction="""Review PMO job descriptions for:                                                      
-    1. Alignment with lean/agile principles                                                              
-    2. Clear role differentiation                                                                        
-    3. Progressive responsibility levels                                                                 
-    4. Consistent formatting and structure                                                               
-    5. Telecommunications industry relevance                                                             
-    6. Emphasis on practical experience over formal requirements                                         
-                                                                                                        
-    Provide specific feedback for improvements.""",
-    model="gpt-4.1",
-)
-@fast.agent(
-    name="file_handler",
-    instruction="""Save the finalized job descriptions as individual Markdown files.                     
-    Use consistent naming like 'pmo_director.md', 'pmo_manager.md' etc.""",
-    servers=["filesystem"],
-    use_history=False,
-)
-@fast.evaluator_optimizer(
-    name="job_description_writer",
-    generator="content_generator",
-    evaluator="consistency_checker",
-    min_rating="EXCELLENT",
-    max_refinements=2,
-)
-async def main() -> None:
-    async with fast.run() as agent:
-        roles = [
-            "PMO Director",
-            "Portfolio Manager",
-            "Senior Program Manager",
-            "Project Manager",
-            "PMO Analyst",
-            "Project Coordinator",
-        ]
-
-        # Pre-initialize the file_handler to establish a persistent connection
-        await agent.file_handler("Test connection to filesystem")
-
-        for role in roles:
-            # Generate and refine job description
-            description = await agent.job_description_writer(
-                f"Create job description for {role} role"
-            )
-            await agent.file_handler(f"Save this job description: {description}")
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

mcp_agent/resources/examples/internal/prompt_category.py

@@ -1,21 +0,0 @@
-from mcp.server.fastmcp import FastMCP
-from mcp.server.fastmcp.prompts.base import AssistantMessage, UserMessage
-
-mcp = FastMCP("MCP Root Tester")
-
-
-@mcp.prompt(name="category_prompt", description="set up the category protocol")
-def category_prompt():
-    return [
-        UserMessage("Cat"),
-        AssistantMessage("animal"),
-        UserMessage("dog"),
-        AssistantMessage("animal"),
-        UserMessage("quartz"),
-        AssistantMessage("mineral"),
-        #        UserMessage("the sun"),
-    ]
-
-
-if __name__ == "__main__":
-    mcp.run()

mcp_agent/resources/examples/internal/prompt_sizing.py

@@ -1,51 +0,0 @@
-from mcp.server.fastmcp import FastMCP
-from mcp.server.fastmcp.prompts.base import AssistantMessage, UserMessage
-from pydantic import Field
-
-mcp = FastMCP("MCP Prompt Tester")
-
-
-@mcp.prompt(name="sizing_prompt", description="set up the sizing protocol")
-def sizing_prompt():
-    return [
-        UserMessage("What is the size of the moon?"),
-        AssistantMessage("OBJECT: MOON\nSIZE: 3,474.8\nUNITS: KM\nTYPE: MINERAL"),
-        UserMessage("What is the size of the Earth?"),
-        AssistantMessage("OBJECT: EARTH\nSIZE: 12,742\nUNITS: KM\nTYPE: MINERAL"),
-        UserMessage("A tiger"),
-        AssistantMessage("OBJECT: TIGER\nSIZE: 1.2\nUNITS: M\nTYPE: ANIMAL"),
-        UserMessage("Domestic Cat"),
-    ]
-
-
-@mcp.prompt(
-    name="sizing_prompt_units",
-    description="set up the sizing protocol with metric or imperial units",
-)
-def sizing_prompt_units(
-    metric: bool = Field(description="Set to True for Metric, False for Imperial", default=True),
-):
-    if metric:
-        return [
-            UserMessage("What is the size of the moon?"),
-            AssistantMessage("OBJECT: MOON\nSIZE: 3,474.8\nUNITS: KM\nTYPE: MINERAL"),
-            UserMessage("What is the size of the Earth?"),
-            AssistantMessage("OBJECT: EARTH\nSIZE: 12,742\nUNITS: KM\nTYPE: MINERAL"),
-            UserMessage("A tiger"),
-            AssistantMessage("OBJECT: TIGER\nSIZE: 1.2\nUNITS: M\nTYPE: ANIMAL"),
-            UserMessage("Domestic Cat"),
-        ]
-    else:
-        return [
-            UserMessage("What is the size of the moon?"),
-            AssistantMessage("OBJECT: MOON\nSIZE: 2,159.1\nUNITS: MI\nTYPE: MINERAL"),
-            UserMessage("What is the size of the Earth?"),
-            AssistantMessage("OBJECT: EARTH\nSIZE: 7,918\nUNITS: MI\nTYPE: MINERAL"),
-            UserMessage("A tiger"),
-            AssistantMessage("OBJECT: TIGER\nSIZE: 3.9\nUNITS: FT\nTYPE: ANIMAL"),
-            UserMessage("Domestic Cat"),
-        ]
-
-
-if __name__ == "__main__":
-    mcp.run()

mcp_agent/resources/examples/internal/simple.txt

@@ -1,2 +0,0 @@
-hello, world
-

mcp_agent/resources/examples/internal/sizer.py

@@ -1,20 +0,0 @@
-import asyncio
-
-from mcp_agent.core.fastagent import FastAgent
-
-fast = FastAgent("Sizer Prompt Test")
-
-
-@fast.agent(
-    "sizer",
-    "given an object return its size",
-    servers=["sizer", "category"],
-    use_history=True,
-)
-async def main() -> None:
-    async with fast.run() as agent:
-        await agent()
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

mcp_agent/resources/examples/internal/social.py

@@ -1,67 +0,0 @@
-import asyncio
-
-from mcp_agent.core.fastagent import FastAgent
-
-# Create the application
-fast = FastAgent("Social Media Manager")
-
-
-@fast.agent(
-    "url_fetcher",
-    "Given a URL, provide a complete and comprehensive summary",
-    servers=["fetch"],
-)
-@fast.agent(
-    "post_author",
-    """
-    Write a 280 character social media post for any given text. 
-    Respond only with the post, never use hashtags.
-    """,
-)
-@fast.agent("translate_fr", "Translate the text to French.")
-@fast.agent("translate_de", "Translate the text to German.")
-@fast.agent(
-    "review",
-    """
-    Cleanly format the original content and translations for  review by a Social Media manager.
-    Highlight any cultural sensitivities.
-    """,
-    model="sonnet",
-)
-@fast.parallel(
-    "translated_plan",
-    fan_out=["translate_fr", "translate_de"],
-)
-@fast.agent(
-    "human_review_and_post",
-    """
-- You can send a social media post by saving it to a file name 'post-<lang>.md'.
-- NEVER POST TO SOCIAL MEDIA UNLESS THE HUMAN HAS REVIEWED AND APPROVED.
-
-Present the Social Media report to the Human, and then provide direct actionable questions to assist
-the Human in posting the content. 
-
-You are being connected to a Human now, the first message you receive will be a
-Social Media report ready to review with the Human.
-
-""",
-    human_input=True,
-    servers=["filesystem"],
-)
-@fast.chain(
-    "post_writer",
-    sequence=[
-        "url_fetcher",
-        "post_author",
-        "translated_plan",
-        "human_review_and_post",
-    ],
-)
-async def main() -> None:
-    async with fast.run() as agent:
-        # using chain workflow
-        await agent.post_writer.prompt()
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

mcp_agent/resources/examples/prompting/__init__.py

@@ -1,3 +0,0 @@
-"""
-Prompting examples package for MCP Agent.
-"""

mcp_agent/resources/examples/prompting/delimited_prompt.txt

@@ -1,14 +0,0 @@
----USER
-I want to learn about {{topic}}.
-
-Can you tell me about it in detail?
-
----ASSISTANT
-I'd be happy to tell you about {{topic}}!
-
-Here are some key facts about {{topic}}:
-1. It's very interesting
-2. It has a rich history
-3. Many people study it
-
-Would you like me to elaborate on any specific aspect?

mcp_agent/resources/examples/prompting/fastagent.config.yaml

@@ -1,43 +0,0 @@
-# FastAgent Configuration File
-
-# Default Model Configuration:
-#
-# Takes format:
-#   <provider>.<model_string>.<reasoning_effort?> (e.g. anthropic.claude-3-5-sonnet-20241022 or openai.o3-mini.low)
-# Accepts aliases for Anthropic Models: haiku, haiku3, sonnet, sonnet35, opus, opus3
-# and OpenAI Models: gpt-4.1-mini, gpt-4.1, o1, o1-mini, o3-mini
-#
-# If not specified, defaults to "haiku".
-# Can be overriden with a command line switch --model=<model>, or within the Agent constructor.
-
-default_model: haiku
-
-# Logging and Console Configuration:
-logger:
-  # level: "debug" | "info" | "warning" | "error"
-  # type: "none" | "console" | "file" | "http"
-  # path: "/path/to/logfile.jsonl"
-  type: file
-  level: error
-  # Switch the progress display on or off
-  progress_display: true
-
-  # Show chat User/Assistant messages on the console
-  show_chat: true
-  # Show tool calls on the console
-  show_tools: true
-  # Truncate long tool responses on the console
-  truncate_tools: true
-
-# MCP Servers
-mcp:
-  servers:
-    prompts:
-      command: "prompt-server"
-      args: ["sizing.md", "resource.md", "resource-exe.md", "pdf_prompt.md"]
-    hfspace:
-      command: "npx"
-      args: ["@llmindset/mcp-hfspace"]
-    image:
-      command: "uv"
-      args: ["run", "image_server.py"]

mcp_agent/resources/examples/prompting/image_server.py

@@ -1,52 +0,0 @@
-#!/usr/bin/env python3
-"""
-Simple MCP server that responds to tool calls with text and image content.
-"""
-
-import logging
-from pathlib import Path
-
-from mcp.server.fastmcp import Context, FastMCP, Image
-from mcp.types import ImageContent, TextContent
-
-# Configure logging
-logging.basicConfig(level=logging.INFO)
-logger = logging.getLogger(__name__)
-
-# Create the FastMCP server
-app = FastMCP(name="ImageToolServer", debug=True)
-
-
-@app.tool(name="get_image", description="Returns an image with a descriptive text")
-async def get_image(
-    image_name: str = "default", ctx: Context = None
-) -> list[TextContent | ImageContent]:
-    """
-    Returns an image file along with a descriptive text.
-
-    Args:
-        image_name: Name of the image to return (default just returns image.jpg)
-
-    Returns:
-        A list containing a text message and the requested image
-    """
-    try:
-        # Read the image file and convert to base64
-        # Create the response with text and image
-        return [
-            TextContent(type="text", text="Here's your image:"),
-            Image(path="image.jpg").to_image_content(),
-        ]
-    except Exception as e:
-        logger.exception(f"Error processing image: {e}")
-        return [TextContent(type="text", text=f"Error processing image: {str(e)}")]
-
-
-if __name__ == "__main__":
-    # Check if the default image exists
-    if not Path("image.jpg").exists():
-        logger.warning("Default image file 'image.jpg' not found in the current directory")
-        logger.warning("Please add an image file named 'image.jpg' to the current directory")
-
-    # Run the server using stdio transport
-    app.run(transport="stdio")

mcp_agent/resources/examples/prompting/prompt1.txt

@@ -1,6 +0,0 @@
-Hello, World.
-
-This is {{blah}} foo
-
-This is {{terrible}} etc.
-

mcp_agent/resources/examples/prompting/work_with_image.py

@@ -1,19 +0,0 @@
-import asyncio
-from pathlib import Path
-
-from mcp_agent.core.fastagent import FastAgent
-from mcp_agent.core.prompt import Prompt
-
-# Create the application
-fast = FastAgent("FastAgent Example")
-
-
-# Define the agent
-@fast.agent("agent", instruction="You are a helpful AI Agent", servers=["prompts"])
-async def main() -> None:
-    async with fast.run() as agent:
-        await agent.agent.generate([Prompt.user("What's in this image?", Path("image.png"))])
-
-
-if __name__ == "__main__":
-    asyncio.run(main())