All templates By NappAI team Outlook Inbox Mail Summarizer
Flow that summarizes unread emails in Outlook's inbox.
graph TD
%%{init: {'theme': 'mc','layout': 'elk'}}%%
ParseData-w4yzj[<div><img src="/_astro/braces.Djq0PW4_.svg" style="height: 20px !important;width: 20px !important"/></div>Parse Data]
style ParseData-w4yzj stroke:#a170ff
Prompt-8bk4w[<div><img src="/_astro/square-terminal.BMOXc-nZ.svg" style="height: 20px !important;width: 20px !important"/></div>Prompt]
style Prompt-8bk4w stroke:#a170ff
ChatOutput-1d6fq[<div><img src="/_astro/messages-square.BaSDmT6g.svg" style="height: 20px !important;width: 20px !important"/></div>Chat Output]
style ChatOutput-1d6fq stroke:#a170ff
OpenAIModel-1ojkm[<div><img src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAI]
style OpenAIModel-1ojkm stroke:#a170ff
OutlookEmailComponent-co5vi[Email]
style OutlookEmailComponent-co5vi stroke:#a170ff
ParseData-w4yzj -.- Prompt-8bk4w
linkStyle 0 stroke:#a170ff
Prompt-8bk4w -.- OpenAIModel-1ojkm
linkStyle 1 stroke:#a170ff
OpenAIModel-1ojkm -.- ChatOutput-1d6fq
linkStyle 2 stroke:#a170ff
OutlookEmailComponent-co5vi -.- ParseData-w4yzj
linkStyle 3 stroke:#a170ff
Outlook Inbox Mail Summarizer
🧩 Overview
The Outlook Inbox Mail Summarizer automates the extraction and summarization of unread emails from a Microsoft Outlook inbox. It pulls the email content, converts it to a plain‑text format, feeds that text into a language‑model prompt, and finally displays a concise, category‑based summary in the chat interface. This workflow streamlines email triage by providing instant, actionable insights without manual reading.
⚙️ Main Features
- Seamless Outlook integration that retrieves unread messages from the user’s inbox.
- Structured text extraction of email fields (sender, subject, body, attachments) using a template.
- Dynamic prompt generation that formats the collected emails for the language model.
- Advanced summarization by an OpenAI model that groups similar messages, highlights key themes, and flags urgent items.
- Chat‑style output that presents the summary in a user‑friendly, conversational format.
🔄 Workflow Steps
| Component Name | Role in the Workflow | Key Inputs | Key Outputs |
|---|---|---|---|
| Outlook Email | Retrieves unread emails from the Outlook inbox. | None (uses stored Outlook credentials). | Data – a collection of email objects. |
| Parse Data | Transforms email objects into plain‑text using a user‑defined template. | Data – the email collection. | Text – formatted email content. |
| Prompt | Builds a prompt string that instructs the LLM to summarize the emails. | Text – formatted email content. | Prompt – a Message ready for the LLM. |
| OpenAI Model | Generates a summarized, category‑based response to the prompt. | Prompt – the instruction and email data. | Text – the final summary. |
| Chat Output | Displays the LLM’s summary in the chat interface. | Text – the summary output. | Message – the displayed chat response. |
🧠Notes
- Authentication: The workflow requires valid Microsoft Outlook and OpenAI credentials; credentials are stored securely and not exposed in the flow.
- Attachment handling: Attachments are fetched only when
Fetch Attachmentsis enabled; text extraction from attachments is optional and limited to common formats (PDF, DOCX). - Batch processing: Mapping mode is disabled by default; the workflow processes the inbox as a single batch.
- Token limits: The OpenAI model may truncate extremely long email collections; consider reducing
max_countif summaries are truncated. - Privacy: Email content is passed to an external LLM; ensure compliance with organizational data‑handling policies before deployment.