docs: fix readmes for examples (#49)

Author: aduminuc

examples/README.md

@@ -1,32 +1,56 @@
-# Email Reviewer (LlamaIndex)
+# Email Reviewer
 
-A simple agent written in LlamaIndex (using a `Workflow`) in charge of reviewing and correcting an email
-
-## Agent Input
-- `email`: (str) Email to review
-- `audience`: (str) one of `general`, `technical`, `business`, `academic`
-
-## Agent Output
-- `correct`: (bool) If given email does not contain writing errors and it targets the audience correctly.
-- `corrected_email`: (opt, str) The corrected email
+The Email Reviewer Agent is a demonstration AI agent developed with Llama-index, designed to review email to address a specific target audience.
 
 ## Requirements
 
- - LlamaIndex: https://docs.llamaindex.ai/en/stable/getting_started/installation/
-
-## Agent local deployment
-
-1) Copy and adapt `.env`: `cp .env.example .env`
-1) Install llama-deploy: `pip install llama_deploy`
-1) Run apiserver:
-    - `python -m llama_deploy.apiserver`
-    OR
-    - (docker): `docker run -p 4501:4501 -v .:/opt/quickstart -w /opt/quickstart llamaindex/llama-deploy:main`
-1) Create deployment config from template: `sed "s|\${PWD}|$(pwd)|g" "email_reviewer.tmpl.yaml" > "email_reviewer.yaml"`
-1) Deploy the workflow: `llamactl deploy email_reviewer.yaml`
+- LlamaIndex: https://docs.llamaindex.ai/en/stable/getting_started/installation/
+- **Azure OpenAI API Key**
+- **Python 3.9 or higher**
+- **Poetry**
+
+## Running the Email Reviewer agent using LlamaDeploy
+
+
+* Copy and adapt `.env`: 
+    ```
+    cp .env.example .env
+    ```
+*  Install LlamaDeploy: 
+    ```
+    pip install llama_deploy
+    ```
+*  Run LlamaDeploy:
+    ```
+    python -m llama_deploy.apiserver`
+    ```
+    or
+    ```
+    docker run -p 4501:4501 -v .:/opt/quickstart -w /opt/quickstart llamaindex/llama-deploy:main
+    ```
+* Create deployment config from template: 
+    ```
+    sed "s|\${PWD}|$(pwd)|g" "email_reviewer.tmpl.yaml" > "email_reviewer.yaml"
+    ```
+
+*  Deploy the workflow: 
+    ```
+    llamactl deploy email_reviewer.yaml
+    ```
+
+## Running the Email Agent using Agent Workflow Server
+
+* Copy and adapt `.env`: `cp .env.example .env`
+* Make sure that the workflow server manager cli (`wfsm`) is added to your path
+* Start the workflow server
+    ```
+    cd deploy;
+    wfsm deploy -m ./email_reviewer.json -e .mailcomposer.env
+    ```
 
 ## Test agent
 
 You can use `usage_example.py` to use the Llama Client SDK to call the agent with given inputs (see source code):
 `python usage_example.py `
 
+

examples/email_reviewer/README.md

@@ -1,22 +1,30 @@
-# Compose Email Agent
+# Email Composer Agent
 
-The Compose Email Agent is built using LangGraph to facilitate interactive email composition.
-It collects email details by continuously interacting with the user until tha latter confirm that the email is ready, then provides the finalized email text.
-
-## Features
-
-- Utilizes Azure OpenAI to guide email composition.
-- Demonstrates the use of LangGraph for managing conversational state and flow.
-- Provides a LangGraph server setup for interacting via ACP protocol APIs.
+The Email Composer Agent is a demonstration AI agent developed with LangGraph, designed to assist in the composition of emails. 
+It gathers necessary email details through ongoing user engagement in a chat format until the user affirms that the email is complete. Subsequently, it delivers the finalized email text.
 
 ## Prerequisites
 
 Before running the application, ensure you have the following:
 
-- **Azure OpenAI API Key**: Set up as an environment variable.
-- **Poetry**: A tool for dependency management and packaging in Python.
+- **Azure OpenAI API Key**
+- **Python 3.9 or higher**
+- **Poetry**
 
+## Running the Email Agent directly
+* Install dependencies
+    ```
+    poetry install
+    ```
 
+* Run the agent
+    ```
+    export AZURE_OPENAI_MODEL=gpt-4o-mini
+    export AZURE_OPENAI_API_KEY=***YOUR_OPENAI_API_KEY***
+    export AZURE_OPENAI_ENDPOINT=***YOUR_OPENAI_ENDPOINT***
+    export OPENAI_API_VERSION=2024-07-01-preview 
+    poetry run python main.py 
+    ```
 
 ## Running the Email Agent using Agent Workflow Server
 
@@ -25,7 +33,7 @@ Before running the application, ensure you have the following:
     echo "
     AZURE_OPENAI_MODEL=gpt-4o-mini
     AZURE_OPENAI_API_KEY=***YOUR_OPENAI_API_KEY***
-    AZURE_OPENAI_ENDPOINT=https://smith-project-agents.openai.azure.com
+    AZURE_OPENAI_ENDPOINT=***YOUR_OPENAI_ENDPOINT***
     OPENAI_API_VERSION=2024-07-01-preview 
     API_HOST=0.0.0.0
     " > deploy/.mailcomposer.env

examples/mailcomposer/README.md

@@ -1,5 +1,68 @@
-poetry run acp generate-agent-models examples/marketing-campaign/deploy/sendgrid.json --output-dir examples/marketing-campaign/src/marketing_campaign --model-file-name sendgrid.py
+# Marketing Campaign Manager
+
+The Marketing Campaign Manager is a demonstration AI agent developed with LangGraph, designed to assist in composing and sending emails for a marketing campaign.
+
+It performs the following actions:
+* It gathers necessary campaign details from the user through a chat.
+* Compose an email leveraging the [Email Composer Agent](../mailcomposer/) as a remote ACP agent.
+* It leverages the [IO Mapper Agent](https://github.com/agntcy/iomapper-agnt) to adapt Email Composer Agent output to Email Reviewer Agent.
+* Reviews the email leveraging the [Email Reviewer Agent](../email_reviewer/) as a remote ACP agent.
+* Send the email to the configured recipient through Twilio sendgrid leveraging the [API Bridge Agent](https://github.com/agntcy/api-bridge-agnt)
+
+## Prerequisites
+
+Before running the application, ensure you have the following:
+
+- **Azure OpenAI API Key**
+- **Python 3.9 or higher**
+- **Poetry**
+
+## Running the Marketing Campaign Manager
+
+### Preliminary step: Deploying API Bridge Agent
+
+Run the API Bridge Agent configured with sendgrid. See [instructions](https://docs.agntcy.org/pages/syntactic_sdk/api_bridge_agent.html#an-example-with-sendgrid-api)
+
+
+### Running the agent directly
+* Deploy agentic dependencies first:
+    * Run Email Composer Agent through the Agent Workflow Server. See [instructions](../mailcomposer/README.md#running-the-email-agent-using-agent-workflow-server).
+    * Run Email Reviewer Agent through the Agent Workflow Server. See [instructions](../email_reviewer/README.md#running-the-email-agent-using-agent-workflow-server).
+* Install python dependencies  
+    ```
+    poetry install
+    ```
+* Run the agent
+    ```
+    export AZURE_OPENAI_MODEL=gpt-4o-mini
+    export AZURE_OPENAI_API_KEY=***YOUR_OPENAI_API_KEY***
+    export AZURE_OPENAI_ENDPOINT=***YOUR_OPENAI_ENDPOINT***
+    export OPENAI_API_VERSION=2024-07-01-preview 
+    export SENDGRID_API_KEY=***YOUR_SENDGRID_API_KEY***
+    export MAILCOMPOSER_HOST=***URL TO EMAIL COMPOSER WORKFLOW SERVER e.g. http://0.0.0.0:10100***
+    export EMAIL_REVIEWER_HOST=***URL TO EMAIL REVIEWER WORKFLOW SERVER e.g. http://0.0.0.0:10100***
+    export MAILCOMPOSER_AGENT_ID==***EMAIL COMPOSER ID**
+    export EMAIL_REVIEWER_AGENT_ID=***EMAIL REVIEWER ID**
+    export SENDGRID_HOST=***URL TO API BRIDGE AGENT*** 
+    
+    poetry run python src/marketing_campaign/__main__.py
+    ```
+
+## Running the Email Agent using Agent Workflow Server
+
+* Create an environment variable file with the variables described above
+    ```commandline
+    echo "
+    AZURE_OPENAI_MODEL=gpt-4o-mini
+    ...
+    ...
+    " > deploy/.mcm.env
+    ```
+* Make sure that the workflow server manager cli (`wfsm`) is added to your path
+* Start the workflow server
+    ```
+    cd deploy;
+    wfsm deploy -m ./mailcomposer.json -e ./.mcm.env
+    ```
 
-poetry run acp generate-agent-models examples/mailcomposer/deploy/mailcomposer.json --output-dir examples/marketing-campaign/src/marketing_campaign --model-file-name mailcomposer.py
 
-poetry run acp generate-agent-models examples/email_reviewer/deploy/email_reviewer.json --output-dir examples/marketing-campaign/src/marketing_campaign --model-file-name email_reviewer.py