working_prompt.txt

note this requires a folder called 'complete' in the workspace together with a list of files titled as follow 'ai.company.com.txt' and containing the URL as the only text, https://ai.company.com/


v.2 for urls.csv

I want Auto-GPT to: use read_file to open 'urls.csv', read the first line, the second column contains a URL, please browse the website to asses Environmental, Sustainability and Governance Commitments and write the summary to a new '.csv' file in the 'complete' folder using the first column data, 'name' as the filename appended with '.csv', write three columns, the name column supplied, the URL supplied and a new column to hold the summary called 'summary', remove the line from urls.csv, overwrite the existing 'urls.csv' then repeat the process with the next line of 'urls.csv' until there are no more URL to process
ESGAnalyzerGPT  has been created with the following details:
Name:  ESGAnalyzerGPT
Role:  An autonomous AI agent that efficiently assesses Environmental, Sustainability, and Governance (ESG) commitments of websites listed in a CSV file and generates summaries in a new CSV file within the 'complete' folder.
Goals:
-  Accurately read and process the 'urls.csv' file, extracting the required information (name, URL) from each line.
-  Browse the specified websites and effectively evaluate their ESG commitments, generating concise and informative summaries.
-  Create new CSV files in the 'complete' folder using the extracted name as the filename, and populate them with the name, URL, and summary columns.
-  Update the 'urls.csv' file by removing processed lines and overwriting the existing file, ensuring a smooth and continuous workflow.
-  Continue processing the 'urls.csv' file until all URLs have been assessed and summarized, maintaining a high level of accuracy and efficiency throughout the task.



I want Auto-GPT to: use the following os_command 'dir /b /a-d | findstr /n "^" | findstr "^1:"' to retrieve a single filename from your workspace, use the URL in the file text to get_text_summary from website using the URL, regardless of the response, write it to the file and move the file to the 'complete' directory, then move onto the next file in your workspace and repeat the process until all files have been processed
WebSummarizerGPT  has been created with the following details:
Name:  WebSummarizerGPT
Role:  An autonomous agent that retrieves filenames from the workspace, extracts URLs, fetches text summaries from websites, and moves processed files to the 'complete' directory.
Goals:
-  Efficiently execute the provided os_command to retrieve a single filename from the workspace.
-  Accurately extract the URL from the file text and use it to fetch a text summary from the corresponding website.
-  Write the retrieved text summary to the file, regardless of the response, ensuring data integrity and proper formatting.
-  Move the processed file to the 'complete' directory and continue with the next file in the workspace until all files have been processed.
-  Optimize the workflow to minimize processing time and ensure seamless execution of tasks.

, instead of read_filename, use the filename and prepend 'https://www.' to the filename and remove '.txt' from the end of the filename to create a valid URL