Skip to content

Commit fceaab9

Browse files
areibmancursoragentalex
authored
Cursor/improve agno tool integrations py output d88d (#1134)
* Refactor tool integration example to self-contained demo with enhanced features Co-authored-by: alex <[email protected]> * fixed examples --------- Co-authored-by: Cursor Agent <[email protected]> Co-authored-by: alex <[email protected]>
1 parent 424ea6c commit fceaab9

File tree

2 files changed

+456
-89
lines changed

2 files changed

+456
-89
lines changed

‎examples/agno/agno_tool_integrations.ipynb‎

Lines changed: 338 additions & 24 deletions
Original file line numberDiff line numberDiff line change
@@ -1,28 +1,342 @@
11
{
2-
"cells": [
3-
{
4-
"cell_type": "markdown",
5-
"id": "813b80b7",
6-
"metadata": {},
7-
"source": [
8-
"\n",
9-
"# Tool Integration with RAG (Retrieval-Augmented Generation) in Agno\n",
10-
"\n",
11-
"This example demonstrates how to enhance Agno agents with RAG capabilities, allowing them to access and reason over external knowledge bases for more accurate and source-backed responses.\n",
12-
"\n",
13-
"**Overview**\n",
14-
"\n",
15-
"This example shows how to integrate RAG with Agno agents where we:\n",
16-
"\n",
17-
"1. **Set up a knowledge base** with documents, URLs, and other external sources\n",
18-
"2. **Configure vector databases** (like Pinecone, Weaviate, or ChromaDB) for efficient semantic search\n",
19-
"3. **Implement retrieval** using embeddings and reranking for accurate information access\n",
20-
"4. **Create RAG-enabled agents** that can search, retrieve, and reason over the knowledge base\n",
21-
"\n",
22-
"By using RAG, agents can provide responses backed by external sources rather than relying solely on their training data, significantly improving accuracy and verifiability of their outputs.\n",
23-
"\n",
24-
"RAG enables agents to access and reason over large knowledge bases,\n"
25-
]
2+
"cells": [
3+
{
4+
"cell_type": "markdown",
5+
"metadata": {},
6+
"source": [
7+
"# Tool Integration Example with Agno\n",
8+
"\n",
9+
"This example demonstrates how to integrate and use various tools with Agno agents, showing how AgentOps automatically tracks tool usage and agent interactions.\n",
10+
"\n",
11+
"## Overview\n",
12+
"\n",
13+
"This example demonstrates:\n",
14+
"\n",
15+
"1. **Using built-in Agno tools** like GoogleSearch, DuckDuckGo, and Arxiv\n",
16+
"2. **Creating agents with tools** and seeing how they use them\n",
17+
"3. **Tool execution tracking** with AgentOps\n",
18+
"4. **Combining multiple tools** for comprehensive research\n",
19+
"\n",
20+
"## Key Concepts\n",
21+
"\n",
22+
"### Tool Integration\n",
23+
"Agno provides various built-in tools that agents can use to access external information and services. These tools are automatically invoked by agents based on the task at hand.\n",
24+
"\n",
25+
"### Available Tools\n",
26+
"- **GoogleSearchTools**: Search the web using Google\n",
27+
"- **DuckDuckGoTools**: Alternative web search engine\n",
28+
"- **ArxivTools**: Search academic papers on arXiv\n",
29+
"- **HackerNewsTools**: Access HackerNews content\n",
30+
"- And many more...\n",
31+
"\n",
32+
"### Automatic Tool Selection\n",
33+
"Agents intelligently select which tools to use based on their instructions and the user's query. They can also combine multiple tools to provide comprehensive responses."
34+
]
35+
},
36+
{
37+
"cell_type": "code",
38+
"execution_count": null,
39+
"metadata": {},
40+
"outputs": [],
41+
"source": [
42+
"# Install the required dependencies\n",
43+
"%pip install agentops\n",
44+
"%pip install agno\n",
45+
"%pip install python-dotenv"
46+
]
47+
},
48+
{
49+
"cell_type": "code",
50+
"execution_count": null,
51+
"metadata": {},
52+
"outputs": [],
53+
"source": [
54+
"import os\n",
55+
"from dotenv import load_dotenv\n",
56+
"\n",
57+
"import agentops\n",
58+
"from agno.agent import Agent\n",
59+
"from agno.models.openai import OpenAIChat\n",
60+
"from agno.tools.googlesearch import GoogleSearchTools\n",
61+
"from agno.tools.duckduckgo import DuckDuckGoTools\n",
62+
"from agno.tools.arxiv import ArxivTools"
63+
]
64+
},
65+
{
66+
"cell_type": "code",
67+
"execution_count": null,
68+
"metadata": {},
69+
"outputs": [],
70+
"source": [
71+
"load_dotenv()\n",
72+
"os.environ[\"OPENAI_API_KEY\"] = os.getenv(\"OPENAI_API_KEY\", \"your_openai_api_key_here\")\n",
73+
"os.environ[\"AGENTOPS_API_KEY\"] = os.getenv(\"AGENTOPS_API_KEY\", \"your_agentops_api_key_here\")"
74+
]
75+
},
76+
{
77+
"cell_type": "code",
78+
"execution_count": null,
79+
"metadata": {},
80+
"outputs": [],
81+
"source": [
82+
"agentops.init(auto_start_session=False, tags=[\"agno-tools\", \"tool-integration\", \"demo\"])"
83+
]
84+
},
85+
{
86+
"cell_type": "markdown",
87+
"metadata": {},
88+
"source": [
89+
"## Example 1: Agent with Single Tool\n",
90+
"\n",
91+
"Let's start with a simple example where we create an agent that uses Google Search to find information. This demonstrates:\n",
92+
"- How to create an agent with a specific tool\n",
93+
"- How the agent automatically uses the tool when needed\n",
94+
"- How AgentOps tracks the tool execution"
95+
]
96+
},
97+
{
98+
"cell_type": "code",
99+
"execution_count": null,
100+
"metadata": {},
101+
"outputs": [],
102+
"source": [
103+
"# Start an AgentOps trace for tracking\n",
104+
"tracer = agentops.start_trace(trace_name=\"Example 1: Single Tool Agent\")\n",
105+
"\n",
106+
"try:\n",
107+
" # Create an agent with Google Search tool\n",
108+
" search_agent = Agent(\n",
109+
" name=\"Search Agent\",\n",
110+
" role=\"Research information using Google Search\",\n",
111+
" model=OpenAIChat(id=\"gpt-4o-mini\"),\n",
112+
" tools=[GoogleSearchTools()],\n",
113+
" instructions=\"You are a research assistant. Use Google Search to find accurate, up-to-date information.\"\n",
114+
" )\n",
115+
" \n",
116+
" # Run the agent with a query\n",
117+
" response = search_agent.run(\"What are the latest developments in AI agents?\")\n",
118+
" \n",
119+
" print(\"Search Agent Response:\")\n",
120+
" print(\"-\" * 60)\n",
121+
" print(response.content)\n",
122+
" print(\"-\" * 60)\n",
123+
" \n",
124+
" agentops.end_trace(tracer, end_state=\"Success\")\n",
125+
" \n",
126+
"except Exception as e:\n",
127+
" print(f\"An error occurred: {e}\")\n",
128+
" agentops.end_trace(tracer, end_state=\"Error\")"
129+
]
130+
},
131+
{
132+
"cell_type": "markdown",
133+
"metadata": {},
134+
"source": [
135+
"## Example 2: Agent with Multiple Tools\n",
136+
"\n",
137+
"Now let's create an agent that has access to multiple tools. The agent will intelligently decide which tool(s) to use based on the query. This demonstrates:\n",
138+
"- How to provide multiple tools to an agent\n",
139+
"- How agents select appropriate tools\n",
140+
"- How to combine information from different sources"
141+
]
142+
},
143+
{
144+
"cell_type": "code",
145+
"execution_count": null,
146+
"metadata": {},
147+
"outputs": [],
148+
"source": [
149+
"# Start a new trace for Example 2\n",
150+
"tracer = agentops.start_trace(trace_name=\"Example 2: Multi-Tool Agent\")\n",
151+
"\n",
152+
"try:\n",
153+
" # Create an agent with multiple research tools\n",
154+
" research_agent = Agent(\n",
155+
" name=\"Research Agent\",\n",
156+
" role=\"Comprehensive research using multiple tools\",\n",
157+
" model=OpenAIChat(id=\"gpt-4o-mini\"),\n",
158+
" tools=[\n",
159+
" GoogleSearchTools(),\n",
160+
" ArxivTools(),\n",
161+
" DuckDuckGoTools()\n",
162+
" ],\n",
163+
" instructions=\"\"\"You are a comprehensive research assistant. \n",
164+
" Use Google Search for general information, Arxiv for academic papers, \n",
165+
" and DuckDuckGo as an alternative search engine. \n",
166+
" Provide well-researched, balanced information from multiple sources.\"\"\"\n",
167+
" )\n",
168+
" \n",
169+
" # Run the agent with a complex query\n",
170+
" response = research_agent.run(\n",
171+
" \"Find information about recent advances in tool-use for AI agents. \"\n",
172+
" \"Include both academic research and practical implementations.\"\n",
173+
" )\n",
174+
" \n",
175+
" print(\"Research Agent Response:\")\n",
176+
" print(\"-\" * 60)\n",
177+
" print(response.content)\n",
178+
" print(\"-\" * 60)\n",
179+
" \n",
180+
" agentops.end_trace(tracer, end_state=\"Success\")\n",
181+
" \n",
182+
"except Exception as e:\n",
183+
" print(f\"An error occurred: {e}\")\n",
184+
" agentops.end_trace(tracer, end_state=\"Error\")"
185+
]
186+
},
187+
{
188+
"cell_type": "markdown",
189+
"metadata": {},
190+
"source": [
191+
"## Example 3: Specialized Tool Usage\n",
192+
"\n",
193+
"Let's create an agent specialized in academic research using the Arxiv tool. This demonstrates:\n",
194+
"- How to create domain-specific agents\n",
195+
"- How specialized tools provide focused results\n",
196+
"- The value of tool specialization for specific tasks"
197+
]
198+
},
199+
{
200+
"cell_type": "code",
201+
"execution_count": null,
202+
"metadata": {},
203+
"outputs": [],
204+
"source": [
205+
"# Start a new trace for Example 3\n",
206+
"tracer = agentops.start_trace(trace_name=\"Example 3: Academic Research Agent\")\n",
207+
"\n",
208+
"try:\n",
209+
" # Create an agent specialized in academic research\n",
210+
" academic_agent = Agent(\n",
211+
" name=\"Academic Agent\",\n",
212+
" role=\"Find and summarize academic papers\",\n",
213+
" model=OpenAIChat(id=\"gpt-4o-mini\"),\n",
214+
" tools=[ArxivTools()],\n",
215+
" instructions=\"You are an academic research assistant. Use Arxiv to find relevant papers and provide concise summaries.\"\n",
216+
" )\n",
217+
" \n",
218+
" # Run the agent with an academic query\n",
219+
" response = academic_agent.run(\n",
220+
" \"Find recent papers about tool augmented language models\"\n",
221+
" )\n",
222+
" \n",
223+
" print(\"Academic Agent Response:\")\n",
224+
" print(\"-\" * 60)\n",
225+
" print(response.content)\n",
226+
" print(\"-\" * 60)\n",
227+
" \n",
228+
" agentops.end_trace(tracer, end_state=\"Success\")\n",
229+
" \n",
230+
"except Exception as e:\n",
231+
" print(f\"An error occurred: {e}\")\n",
232+
" agentops.end_trace(tracer, end_state=\"Error\")"
233+
]
234+
},
235+
{
236+
"cell_type": "markdown",
237+
"metadata": {},
238+
"source": [
239+
"## Example 4: Comparing Different Search Tools\n",
240+
"\n",
241+
"Finally, let's create an agent that can compare results from different search engines. This demonstrates:\n",
242+
"- How agents can use multiple tools strategically\n",
243+
"- How to compare and synthesize information from different sources\n",
244+
"- The benefits of having alternative data sources"
245+
]
246+
},
247+
{
248+
"cell_type": "code",
249+
"execution_count": null,
250+
"metadata": {},
251+
"outputs": [],
252+
"source": [
253+
"# Start a new trace for Example 4\n",
254+
"tracer = agentops.start_trace(trace_name=\"Example 4: Tool Comparison Agent\")\n",
255+
"\n",
256+
"try:\n",
257+
" # Create an agent that compares different search tools\n",
258+
" comparison_agent = Agent(\n",
259+
" name=\"Comparison Agent\",\n",
260+
" role=\"Compare results from different search engines\",\n",
261+
" model=OpenAIChat(id=\"gpt-4o-mini\"),\n",
262+
" tools=[\n",
263+
" GoogleSearchTools(),\n",
264+
" DuckDuckGoTools()\n",
265+
" ],\n",
266+
" instructions=\"\"\"Compare search results from Google and DuckDuckGo. \n",
267+
" Note any differences in results, ranking, or information quality.\n",
268+
" Be objective in your comparison.\"\"\"\n",
269+
" )\n",
270+
" \n",
271+
" # Run the agent with a comparison task\n",
272+
" response = comparison_agent.run(\n",
273+
" \"Search for 'AgentOps observability platform' on both search engines and compare the results\"\n",
274+
" )\n",
275+
" \n",
276+
" print(\"Comparison Agent Response:\")\n",
277+
" print(\"-\" * 60)\n",
278+
" print(response.content)\n",
279+
" print(\"-\" * 60)\n",
280+
" \n",
281+
" agentops.end_trace(tracer, end_state=\"Success\")\n",
282+
" \n",
283+
"except Exception as e:\n",
284+
" print(f\"An error occurred: {e}\")\n",
285+
" agentops.end_trace(tracer, end_state=\"Error\")"
286+
]
287+
},
288+
{
289+
"cell_type": "markdown",
290+
"metadata": {},
291+
"source": [
292+
"## Key Takeaways\n",
293+
"\n",
294+
"This notebook demonstrated several important concepts about tool integration in Agno:\n",
295+
"\n",
296+
"1. **Automatic Tool Usage**: Agents automatically determine when and how to use their available tools based on the query.\n",
297+
"\n",
298+
"2. **Tool Selection**: When given multiple tools, agents intelligently select the most appropriate ones for the task.\n",
299+
"\n",
300+
"3. **Information Synthesis**: Agents can combine information from multiple tools to provide comprehensive responses.\n",
301+
"\n",
302+
"4. **AgentOps Tracking**: All tool executions are automatically tracked by AgentOps, providing observability into:\n",
303+
" - Which tools were used\n",
304+
" - What parameters were passed\n",
305+
" - Tool execution times\n",
306+
" - Success/failure rates\n",
307+
"\n",
308+
"5. **Specialized vs. General Agents**: You can create both specialized agents (with specific tools) and general-purpose agents (with multiple tools).\n",
309+
"\n",
310+
"## Viewing Results in AgentOps\n",
311+
"\n",
312+
"After running these examples, you can visit [AgentOps Dashboard](https://app.agentops.ai/) to see:\n",
313+
"- Detailed traces of each agent execution\n",
314+
"- Tool usage patterns\n",
315+
"- Performance metrics\n",
316+
"- Error tracking and debugging information\n",
317+
"\n",
318+
"Each trace URL printed during execution provides direct access to that specific session's details."
319+
]
320+
}
321+
],
322+
"metadata": {
323+
"kernelspec": {
324+
"display_name": "Python 3",
325+
"language": "python",
326+
"name": "python3"
327+
},
328+
"language_info": {
329+
"codemirror_mode": {
330+
"name": "ipython",
331+
"version": 3
332+
},
333+
"file_extension": ".py",
334+
"mimetype": "text/x-python",
335+
"name": "python",
336+
"nbconvert_exporter": "python",
337+
"pygments_lexer": "ipython3",
338+
"version": "3.9.0"
339+
}
26340
},
27341
{
28342
"cell_type": "code",

0 commit comments

Comments
 (0)