Ap/add websearch component

[alimosaed]

Changes Made

• Implemented web search and scraping agents for Google, DuckDuckGo, and Bing.
• Created stand-alone workflows for each of the new agents.
• Developed a comprehensive flow that routes requests to either web search or LLM based on the request body.
• Introduced a filter to trim the LLM output by removing unnecessary data from the JSON object.
• Enabled the response_format tag in openai_chat_model_base.py to force OpenAI to generate a trimmed JSON object.
• Added 'unicode_escape' encoding to the broker_base.py to generate results without '\n'.

Testing

• Build the Solace AI Connector. Navigate to the root of the project and run

pip install .

• Run each YAML files by

solace-ai-connector <name>.yaml

• The required inputs per use case have been documented on top of each YAML file.

Documentation

• The required packages have been documented on Readme.md files.
• The required inputs per use case have been documented on top of each YAML file.

[gitstream-cm]

Please mark whether you used Copilot to assist coding in this PR

• Copilot Assisted

[cyrus2281]

Added some comments

[cyrus2281]

Looks good.

Please just run the doc generator command to update the documents.

Just to clarify, the parser does not support tag extraction and then parsing. it only takes in the whole passed in input

[cyrus2281]

Run make gen-docs

[alimosaed]

• Built documents
• Updated the error messages

[sonarqube-solacecloud]

SonarQube Quality Gate

0.0% Security Hotspots Reviewed on New Code (is less than 100%)

See analysis details on SonarQube