agent-bootstrap-kit 1.0.0__tar.gz

agent_bootstrap_kit-1.0.0/PKG-INFO

@@ -0,0 +1,78 @@
+Metadata-Version: 2.4
+Name: agent-bootstrap-kit
+Version: 1.0.0
+Summary: OpenAI agent bootstrap scripts for assessments
+License: MIT
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: openai>=1.0.0
+Requires-Dist: python-dotenv>=1.0.0
+Requires-Dist: pyautogen>=0.2.0
+Requires-Dist: semantic-kernel>=1.0.0
+Requires-Dist: azure-ai-projects>=1.0.0
+Requires-Dist: azure-identity>=1.0.0
+
+# agent-bootstrap-kit
+
+Bootstrap scripts for building OpenAI agents — multi-turn chat, tool use, AutoGen, Semantic Kernel, and Azure AI Agent Service.
+
+## Install
+
+```bash
+pip install agent-bootstrap-kit
+```
+
+## Extract scripts to your current folder
+
+```bash
+agent-bootstrap
+```
+
+Or in Python:
+
+```python
+from agent_bootstrap_kit import extract
+extract()
+```
+
+## What gets extracted
+
+```
+plain_openai/
+    1_multiturn_chat.py         # Multi-turn conversation agent
+    2_simple_agent_tools.py     # Agent with tool/function calling
+
+autogen/
+    3_autogen_two_agent.py      # AutoGen two-agent conversation
+    4_autogen_tool_use.py       # AutoGen agent with tools
+
+semantic_kernel/
+    5_sk_basic_agent.py         # Semantic Kernel agent with plugin
+    6_sk_workflow.py            # Semantic Kernel multi-step pipeline
+
+azure_ai_agent/
+    7_azure_agent_thread.py     # Azure AI Agent Service (thread-based)
+
+EXAM_DAY_REFERENCE.py           # Quick reference — which script to use when
+setup.sh                        # Install all dependencies in one command
+.env.template                   # Copy → .env, add your API key
+```
+
+## Quick start after extraction
+
+```bash
+# 1. Add your API key
+cp .env.template .env
+# Edit .env: OPENAI_API_KEY=sk-...
+
+# 2. Install dependencies
+bash setup.sh
+
+# 3. Run a script
+python plain_openai/1_multiturn_chat.py
+```
+
+## Requirements
+
+- Python 3.9+
+- An OpenAI API key

agent_bootstrap_kit-1.0.0/README.md

@@ -0,0 +1,64 @@
+# agent-bootstrap-kit
+
+Bootstrap scripts for building OpenAI agents — multi-turn chat, tool use, AutoGen, Semantic Kernel, and Azure AI Agent Service.
+
+## Install
+
+```bash
+pip install agent-bootstrap-kit
+```
+
+## Extract scripts to your current folder
+
+```bash
+agent-bootstrap
+```
+
+Or in Python:
+
+```python
+from agent_bootstrap_kit import extract
+extract()
+```
+
+## What gets extracted
+
+```
+plain_openai/
+    1_multiturn_chat.py         # Multi-turn conversation agent
+    2_simple_agent_tools.py     # Agent with tool/function calling
+
+autogen/
+    3_autogen_two_agent.py      # AutoGen two-agent conversation
+    4_autogen_tool_use.py       # AutoGen agent with tools
+
+semantic_kernel/
+    5_sk_basic_agent.py         # Semantic Kernel agent with plugin
+    6_sk_workflow.py            # Semantic Kernel multi-step pipeline
+
+azure_ai_agent/
+    7_azure_agent_thread.py     # Azure AI Agent Service (thread-based)
+
+EXAM_DAY_REFERENCE.py           # Quick reference — which script to use when
+setup.sh                        # Install all dependencies in one command
+.env.template                   # Copy → .env, add your API key
+```
+
+## Quick start after extraction
+
+```bash
+# 1. Add your API key
+cp .env.template .env
+# Edit .env: OPENAI_API_KEY=sk-...
+
+# 2. Install dependencies
+bash setup.sh
+
+# 3. Run a script
+python plain_openai/1_multiturn_chat.py
+```
+
+## Requirements
+
+- Python 3.9+
+- An OpenAI API key

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/__init__.py

@@ -0,0 +1,18 @@
+"""
+agent_bootstrap_kit
+===================
+Bootstrap scripts for OpenAI agent assessments.
+
+Usage on the VM:
+    pip install agent-bootstrap-kit
+    agent-bootstrap              # extracts all files to current folder
+    
+    OR in Python:
+    from agent_bootstrap_kit import extract
+    extract()
+"""
+
+from agent_bootstrap_kit.extractor import extract
+
+__version__ = "1.0.0"
+__all__ = ["extract"]

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/cli.py

@@ -0,0 +1,37 @@
+"""
+cli.py
+======
+Entry point for the "agent-bootstrap" command-line tool.
+
+After pip install, this lets you just type:
+    agent-bootstrap
+
+...and it extracts all scripts into your current folder.
+"""
+
+import argparse
+from agent_bootstrap_kit.extractor import extract
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        prog="agent-bootstrap",
+        description="Extract OpenAI agent bootstrap scripts into your current folder"
+    )
+    parser.add_argument(
+        "--dest",
+        default=".",
+        help="Destination folder (default: current directory)"
+    )
+    parser.add_argument(
+        "--version",
+        action="version",
+        version="agent-bootstrap-kit 1.0.0"
+    )
+
+    args = parser.parse_args()
+    extract(destination=args.dest)
+
+
+if __name__ == "__main__":
+    main()

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/extractor.py

@@ -0,0 +1,94 @@
+"""
+extractor.py
+============
+Copies all bootstrap scripts from the installed package
+into the user's current working directory.
+
+This is the magic that makes "pip install → files appear" work.
+"""
+
+import os
+import shutil
+import pathlib
+
+
+def extract(destination: str = ".") -> None:
+    """
+    Extract all bootstrap scripts into the destination folder.
+    
+    Args:
+        destination: Path to copy files into. Defaults to current directory.
+    
+    Usage:
+        from agent_bootstrap_kit import extract
+        extract()           # copies into current folder
+        extract("./mywork") # copies into ./mywork
+    """
+
+    # Find where THIS package is installed on disk
+    # __file__ = .../site-packages/agent_bootstrap_kit/extractor.py
+    package_dir = pathlib.Path(__file__).parent
+    scripts_dir = package_dir / "scripts"
+
+    if not scripts_dir.exists():
+        print(f"ERROR: scripts directory not found at {scripts_dir}")
+        return
+
+    dest = pathlib.Path(destination).resolve()
+    dest.mkdir(parents=True, exist_ok=True)
+
+    print(f"\nExtracting agent bootstrap scripts to: {dest}\n")
+
+    copied = []
+    skipped = []
+
+    # Walk every file inside scripts/ and copy to destination
+    for src_path in scripts_dir.rglob("*"):
+        if src_path.is_dir():
+            continue  # directories are created automatically below
+
+        # Compute relative path from scripts_dir
+        # e.g. scripts/plain_openai/1_multiturn_chat.py → plain_openai/1_multiturn_chat.py
+        relative = src_path.relative_to(scripts_dir)
+        dst_path = dest / relative
+
+        # Create parent directories if needed
+        dst_path.parent.mkdir(parents=True, exist_ok=True)
+
+        # Don't overwrite files the user may have already edited
+        if dst_path.exists():
+            skipped.append(str(relative))
+            continue
+
+        shutil.copy2(src_path, dst_path)
+        copied.append(str(relative))
+
+    # Print a clean summary
+    if copied:
+        print("  ✓ Copied:")
+        for f in sorted(copied):
+            print(f"      {f}")
+
+    if skipped:
+        print("\n  ⚠ Skipped (already exists):")
+        for f in sorted(skipped):
+            print(f"      {f}")
+
+    print(f"\n  Done. {len(copied)} file(s) extracted.\n")
+
+    # Print next steps
+    print("=" * 55)
+    print("NEXT STEPS:")
+    print("=" * 55)
+    print("1. Rename .env.template → .env")
+    print("   Add your key:  OPENAI_API_KEY=sk-...")
+    print("")
+    print("2. Install dependencies:")
+    print("   bash setup.sh")
+    print("   (or: pip install openai pyautogen semantic-kernel python-dotenv)")
+    print("")
+    print("3. Open EXAM_DAY_REFERENCE.py to pick the right script")
+    print("")
+    print("4. Run a script, e.g.:")
+    print("   python plain_openai/1_multiturn_chat.py")
+    print("=" * 55)

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/scripts/EXAM_DAY_REFERENCE.py

@@ -0,0 +1,142 @@
+# ══════════════════════════════════════════════════════════════════
+#  EXAM DAY QUICK REFERENCE — OpenAI Agent Assessment
+#  Keep this open. Match the task to the right script.
+# ══════════════════════════════════════════════════════════════════
+
+
+# ─────────────────────────────────────────────────────────────────
+# STEP 0: FIRST THING — Create your .env file
+# ─────────────────────────────────────────────────────────────────
+# Create a file called ".env" in the same folder as your scripts.
+# Add this line (replace with your real key):
+#
+#   OPENAI_API_KEY=sk-proj-xxxxxxxxxxxxxxxxxxxxxxxx
+#
+# Optional (only if you have Azure):
+#   AZURE_AI_PROJECT_ENDPOINT=https://your-endpoint.api.azureml.ms
+#
+# Then run: bash setup.sh
+# ─────────────────────────────────────────────────────────────────
+
+
+# ─────────────────────────────────────────────────────────────────
+# DECISION TREE: Which script to use?
+# ─────────────────────────────────────────────────────────────────
+#
+#  Task says "build a chatbot" or "multi-turn conversation"
+#  └── Use: plain_openai/1_multiturn_chat.py
+#
+#  Task says "agent with tools" or "function calling" or "agent that can look up data"
+#  └── Use: plain_openai/2_simple_agent_tools.py
+#
+#  Task says "AutoGen" or "multi-agent" or "two agents talking to each other"
+#  └── Basic two agents:  autogen/3_autogen_two_agent.py
+#  └── With tools:        autogen/4_autogen_tool_use.py
+#
+#  Task says "Semantic Kernel" or "SK" or "workflow pipeline"
+#  └── Basic agent:       semantic_kernel/5_sk_basic_agent.py
+#  └── Multi-step workflow: semantic_kernel/6_sk_workflow.py
+#
+#  Task says "Azure AI Agent" or "Azure AI Foundry" or "thread-based agent"
+#  └── Use: azure_ai_agent/7_azure_agent_thread.py
+#  └── (Automatically falls back to plain OpenAI if no Azure endpoint)
+# ─────────────────────────────────────────────────────────────────
+
+
+# ─────────────────────────────────────────────────────────────────
+# COMMON CUSTOMIZATIONS — how to adapt any script
+# ─────────────────────────────────────────────────────────────────
+
+# 1. CHANGE THE PERSONA / SYSTEM PROMPT
+#    Find this line and edit the text:
+#      system_message = "You are a helpful assistant..."
+#    Example: "You are a customer service agent for Acme Corp. Be polite and concise."
+
+# 2. ADD A NEW TOOL (scripts 2 and 4)
+#    a) Write a Python function that does the work
+#    b) Add it to the TOOLS list (JSON schema) with name + description + parameters
+#    c) Add it to TOOL_MAP so it gets called: TOOL_MAP["your_tool_name"] = your_function
+
+# 3. CHANGE THE MODEL
+#    Find MODEL = "gpt-4o" and change to:
+#      "gpt-4-turbo"     → cheaper, slightly older
+#      "gpt-3.5-turbo"   → cheapest, less capable
+#      "gpt-4o-mini"     → fast and cheap
+
+# 4. CHANGE HOW MANY TURNS BEFORE AUTO-STOP (AutoGen scripts)
+#    Find: max_consecutive_auto_reply=5
+#    Increase to 10 or 15 for longer tasks
+
+# 5. MAKE THE AGENT STOP AUTOMATICALLY
+#    In AutoGen: the agent sends "TERMINATE" and UserProxy stops
+#    In plain OpenAI: add a check: if "DONE" in response: break
+# ─────────────────────────────────────────────────────────────────
+
+
+# ─────────────────────────────────────────────────────────────────
+# PACKAGE INSTALL CHEAT SHEET (if setup.sh didn't work)
+# ─────────────────────────────────────────────────────────────────
+#
+#   pip install openai python-dotenv           # always needed
+#   pip install pyautogen                      # for AutoGen scripts
+#   pip install semantic-kernel                # for SK scripts
+#   pip install azure-ai-projects azure-identity  # for Azure script
+# ─────────────────────────────────────────────────────────────────
+
+
+# ─────────────────────────────────────────────────────────────────
+# CORE PATTERNS TO REMEMBER
+# ─────────────────────────────────────────────────────────────────
+
+# MULTI-TURN MEMORY (plain OpenAI) — always append both sides:
+#
+#   messages.append({"role": "user",      "content": user_input})
+#   response = client.chat.completions.create(model=..., messages=messages)
+#   reply = response.choices[0].message.content
+#   messages.append({"role": "assistant", "content": reply})
+
+# TOOL CALL FLOW (plain OpenAI):
+#
+#   1. response = client.chat.completions.create(..., tools=TOOLS, tool_choice="auto")
+#   2. if response.choices[0].finish_reason == "tool_calls":
+#   3.     for each tool_call: run the Python function, append result with role="tool"
+#   4. Call the API again → model produces final answer
+
+# AUTOGEN PATTERN:
+#
+#   assistant = AssistantAgent(name=..., llm_config=..., system_message=...)
+#   user_proxy = UserProxyAgent(name=..., human_input_mode="NEVER", ...)
+#   user_proxy.initiate_chat(assistant, message="Your task here")
+
+# SEMANTIC KERNEL PATTERN:
+#
+#   kernel = Kernel()
+#   kernel.add_service(OpenAIChatCompletion(...))
+#   kernel.add_plugin(MyPlugin(), plugin_name="myplugin")
+#   result = await kernel.invoke(kernel.plugins["myplugin"]["function_name"], args)
+# ─────────────────────────────────────────────────────────────────
+
+
+# ─────────────────────────────────────────────────────────────────
+# TROUBLESHOOTING
+# ─────────────────────────────────────────────────────────────────
+#
+#  Error: "No module named openai"
+#  → Run: pip install openai
+#
+#  Error: "AuthenticationError" or "Invalid API key"
+#  → Check your .env file has OPENAI_API_KEY=sk-...  (no quotes around the key)
+#
+#  Error: "No module named autogen" 
+#  → Run: pip install pyautogen
+#
+#  Error: "No module named semantic_kernel"
+#  → Run: pip install semantic-kernel
+#
+#  Script hangs / no output
+#  → API call is waiting. Check internet connectivity to api.openai.com
+#  → Try: curl https://api.openai.com/v1/models -H "Authorization: Bearer YOUR_KEY"
+#
+#  AutoGen keeps going forever
+#  → Add TERMINATE to the assistant's system_message, or reduce max_consecutive_auto_reply
+# ─────────────────────────────────────────────────────────────────

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/scripts/autogen/3_autogen_two_agent.py

@@ -0,0 +1,75 @@
+"""
+SCRIPT 3: AutoGen — Two-Agent Conversation
+==========================================
+What it does:
+  - Creates two agents: a UserProxy (acts as the human) and an AssistantAgent (the AI)
+  - They converse with each other to complete a task
+  - You just give the initial task — agents handle the back-and-forth
+
+When to use this pattern:
+  - Task automation where you want AI to propose a plan and execute it
+  - Any scenario requiring back-and-forth reasoning
+
+Key concepts:
+  - UserProxyAgent  : represents the human / initiates conversation / can run code
+  - AssistantAgent  : the AI brain that responds and plans
+  - initiate_chat() : kicks off the multi-turn conversation between them
+  - max_consecutive_auto_reply : how many auto-replies before pausing for human input
+"""
+
+import os
+from dotenv import load_dotenv
+import autogen
+
+load_dotenv()
+
+# ── Config: tell AutoGen which LLM to use ───────────────────
+LLM_CONFIG = {
+    "config_list": [
+        {
+            "model": "gpt-4o",
+            "api_key": os.getenv("OPENAI_API_KEY"),
+        }
+    ],
+    "temperature": 0.7,
+}
+
+# ── Create the two agents ────────────────────────────────────
+
+# AssistantAgent: the AI that thinks, plans, writes code, answers
+assistant = autogen.AssistantAgent(
+    name="Assistant",
+    llm_config=LLM_CONFIG,
+    system_message="""
+    You are a helpful AI assistant. 
+    When given a task, break it down into clear steps and complete it.
+    When you have fully answered, end your message with: TERMINATE
+    """
+)
+
+# UserProxyAgent: acts as the human side
+# human_input_mode="NEVER" means it runs fully automatically (no pause for you to type)
+# human_input_mode="TERMINATE" means it runs automatically but stops on TERMINATE keyword
+# human_input_mode="ALWAYS" means it asks for human input at every step
+user_proxy = autogen.UserProxyAgent(
+    name="UserProxy",
+    human_input_mode="NEVER",              # change to "TERMINATE" to review before ending
+    max_consecutive_auto_reply=5,          # stop after 5 auto-replies max
+    is_termination_msg=lambda msg: "TERMINATE" in msg.get("content", ""),
+    code_execution_config={"use_docker": False},  # run code locally, not in Docker
+)
+
+# ── Run the conversation ─────────────────────────────────────
+if __name__ == "__main__":
+    task = """
+    I need a summary of the benefits of multi-agent AI systems.
+    List 5 key advantages with a one-sentence explanation for each.
+    """
+
+    print("Starting AutoGen two-agent conversation...\n")
+    print("=" * 60)
+
+    user_proxy.initiate_chat(
+        recipient=assistant,
+        message=task
+    )

agent_bootstrap_kit-1.0.0/agent_bootstrap_kit/scripts/autogen/4_autogen_tool_use.py

@@ -0,0 +1,106 @@
+"""
+SCRIPT 4: AutoGen — Agent with Tool Use (Function Registration)
+===============================================================
+What it does:
+  - Registers Python functions as tools an AutoGen agent can call
+  - The AssistantAgent decides when to invoke the tool
+  - UserProxy actually executes the tool and returns the result
+  - Models a realistic workflow agent
+
+When to use this pattern:
+  - Workflow automation agent that needs to call services, APIs, or run logic
+  - Research agent, data processing agent, etc.
+
+Key concepts:
+  - @user_proxy.register_for_execution()  : marks the Python function as executable
+  - @assistant.register_for_llm()         : tells the LLM this tool is available
+  - The decorator pattern links definition → LLM awareness → execution
+"""
+
+import os
+import json
+import autogen
+from autogen import AssistantAgent, UserProxyAgent
+from dotenv import load_dotenv
+
+load_dotenv()
+
+LLM_CONFIG = {
+    "config_list": [
+        {
+            "model": "gpt-4o",
+            "api_key": os.getenv("OPENAI_API_KEY"),
+        }
+    ]
+}
+
+# ── Create agents ────────────────────────────────────────────
+assistant = AssistantAgent(
+    name="ToolAssistant",
+    llm_config=LLM_CONFIG,
+    system_message="""
+    You are a helpful assistant with access to tools.
+    Use the tools when you need to fetch data or do calculations.
+    Once you have all the information needed, provide a clear final answer.
+    End with TERMINATE when the task is complete.
+    """
+)
+
+user_proxy = UserProxyAgent(
+    name="UserProxy",
+    human_input_mode="NEVER",
+    max_consecutive_auto_reply=5,
+    is_termination_msg=lambda msg: "TERMINATE" in msg.get("content", ""),
+    code_execution_config=False,  # we're using registered tools, not raw code execution
+)
+
+
+# ── Define and register tools ────────────────────────────────
+# The decorator pattern: one decorator tells the LLM, one handles execution
+
+@user_proxy.register_for_execution()       # UserProxy runs this
+@assistant.register_for_llm(
+    name="get_stock_price",
+    description="Get the current stock price for a given ticker symbol"
+)
+def get_stock_price(ticker: str) -> str:
+    """Simulated stock price lookup."""
+    fake_prices = {
+        "MSFT": "$415.23",
+        "AAPL": "$189.50",
+        "GOOGL": "$175.80",
+        "OPENAI": "Private company, no public price"
+    }
+    return fake_prices.get(ticker.upper(), f"Ticker '{ticker}' not found")
+
+
+@user_proxy.register_for_execution()
+@assistant.register_for_llm(
+    name="get_company_info",
+    description="Get basic information about a company given its name or ticker"
+)
+def get_company_info(company: str) -> str:
+    """Simulated company info lookup."""
+    info = {
+        "MSFT": "Microsoft Corporation. Founded 1975. CEO: Satya Nadella. Sector: Technology.",
+        "AAPL": "Apple Inc. Founded 1976. CEO: Tim Cook. Sector: Technology/Consumer Electronics.",
+        "GOOGL": "Alphabet Inc. Founded 1998. CEO: Sundar Pichai. Sector: Technology/Advertising.",
+    }
+    return info.get(company.upper(), f"No information found for '{company}'")
+
+
+# ── Run the agent with a task that requires tool use ─────────
+if __name__ == "__main__":
+    task = """
+    I'm researching tech stocks. 
+    Please get the current price and company info for Microsoft (MSFT) and Apple (AAPL),
+    then give me a brief comparison of the two companies.
+    """
+
+    print("Starting AutoGen Tool-Use Agent...\n")
+    print("=" * 60)
+
+    user_proxy.initiate_chat(
+        recipient=assistant,
+        message=task
+    )