Fully-featured Agent Recipe (#53)

*add full feature agent workbook + test fixes

Author: rbs333

.github/workflows/test.yml

@@ -7,6 +7,8 @@ on:
   push:
     branches:
       - main
+  schedule:
+    - cron: '0 0 * * 1' # Runs every Monday
 
 jobs:
   test:

README.md

@@ -85,6 +85,7 @@ An estimated 31% of LLM queries are potentially redundant ([source](https://arxi
 | --- | --- |
 [/agents/00_langgraph_redis_agentic_rag.ipynb](python-recipes/agents/00_langgraph_redis_agentic_rag.ipynb) | Notebook to get started with lang-graph and agents |
 [/agents/01_crewai_langgraph_redis.ipynb](python-recipes/agents/01_crewai_langgraph_redis.ipynb) | Notebook to get started with lang-graph and agents |
+[/agents/02_full_featured_agent.ipynb](python-recipes/agents/02_full_featured_agent.ipynb) | Notebook builds full tool calling agent with semantic cache and router |
 
 ### Computer Vision
 | Recipe | Description |

assets/cache_diagram.png

@@ -105,22 +105,22 @@
    "outputs": [],
    "source": [
     "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
-    "from langchain.document_loaders import UnstructuredFileLoader\n",
+    "from langchain_community.document_loaders import PyPDFLoader\n",
     "\n",
     "CHUNK_SIZE = 2500\n",
     "CHUNK_OVERLAP = 0\n",
     "\n",
-    "source_doc = \"resources/nke-10k-2023.pdf\"\n",
+    "# pdf to load\n",
+    "path = 'resources/nke-10k-2023.pdf'\n",
+    "assert os.path.exists(path), f\"File not found: {path}\"\n",
     "\n",
-    "loader = UnstructuredFileLoader(\n",
-    "    source_doc, mode=\"single\", strategy=\"fast\"\n",
-    ")\n",
-    "\n",
-    "text_splitter = RecursiveCharacterTextSplitter(\n",
-    "    chunk_size=CHUNK_SIZE, chunk_overlap=CHUNK_OVERLAP\n",
-    ")\n",
+    "# load and split\n",
+    "loader = PyPDFLoader(path)\n",
+    "pages = loader.load()\n",
+    "text_splitter = RecursiveCharacterTextSplitter(chunk_size=CHUNK_SIZE, chunk_overlap=CHUNK_OVERLAP)\n",
+    "chunks = text_splitter.split_documents(pages)\n",
     "\n",
-    "chunks = loader.load_and_split(text_splitter)"
+    "print(\"Done preprocessing. Created\", len(chunks), \"chunks of the original pdf\", path)"
    ]
   },
   {