Skip to content

README.md

Latest commit

 

History

History
97 lines (73 loc) · 4.93 KB

README.md

File metadata and controls

97 lines (73 loc) · 4.93 KB
Raw

Gmail Newsletter Intel Brief

Overview

This automation pulls useful signals from recent newsletter emails and combines them into one concise digest. It is for people who want the takeaways without reading every newsletter.

How It Works

  1. Looks back over a bounded recent Gmail window and discovers likely recurring newsletter senders.
  2. Reuses the automation memory file to remember confirmed newsletter senders, recurring non-newsletter senders, and compact sender notes from prior runs.
  3. Reads the current in-window issues from the detected senders instead of opportunistically sampling a noisy Gmail category.
  4. Extracts the main items, claims, entities, and links from each issue without trying to summarize the whole email thread universe.
  5. Clusters overlapping items into a short list of recurring themes.
  6. Merges repetitive coverage across newsletters and keeps the clearest, most evidence-backed source links.
  7. Produces one Markdown brief with both coverage accounting and high-signal items.
sequenceDiagram
    participant Agent
    participant Gmail as Gmail Scope
    participant Links as Source Links

    Agent->>Gmail: Read bounded newsletter slice
    Gmail-->>Agent: In-scope messages and metadata
    Agent->>Links: Optionally confirm highest-signal links
    Links-->>Agent: Human-readable article context
    Note over Agent: Read-only, no replies, labels, archive, or deletes
Loading

Prerequisites

  • Gmail access through a configured Google Workspace MCP server or the gws CLI

It can run with no extra configuration. It works best when you later narrow the scope with a label, sender set, or Gmail query that already isolates the newsletters you trust.

Cursor Cloud Usage

  1. Open Cursor Automations.
  2. Name your automation and paste gmail-newsletter-intel-brief.md as the automation prompt.
  3. Add Google Workspace access through a Workspace MCP server or the gws CLI.
  4. Optionally customize the scope or audience notes near the top of the prompt.
  5. Allow memory so the automation can remember confirmed newsletter senders and recurring non-newsletter senders between runs.
  6. Set the schedule or run manually, then save the automation.

Codex App Usage

  1. Set up Gmail access in Codex using a Google Workspace MCP server or the gws CLI.
  2. Click Automation > New Automation.
  3. Name your automation and paste gmail-newsletter-intel-brief.md as the automation prompt.
  4. Optionally customize the scope or audience notes near the top of the prompt.
  5. Let the automation reuse its persistent memory.md between runs for sender allow and ignore hints.
  6. Set schedule or run manually and save the automation.

Claude Code Usage

  1. Add a Google Workspace MCP server to Claude Code if you want MCP-backed runs, or make gws available in the runtime.
  2. Make sure the runtime can read Gmail messages, message metadata, and links from the selected newsletter scope.
  3. Use the prompt as-is for the default behavior, or add a short override note before using /loop or /schedule. For example:
Newsletter Gmail scope: label:newsletters newer_than:7d
Topic priorities: competitors, product launches, tooling, market moves
Audience: product and strategy leads
Output emphasis: keep only items that are new, repeated across sources, or materially actionable
  1. For durable Claude-managed automation outside the current session, use /schedule or create a Routine in claude.ai/code/routines.

Recommended Defaults

Setting Default
Discovery window last 30 days
Brief window last 7 days
Message scope bounded Gmail discovery, then detected recurring newsletter senders
Sender discovery budget up to 150 candidate messages
Full issue reads up to 3 recent issues per detected sender, up to 60 total issue reads
Final theme count up to 6 themes
Final item count up to 12 items
Output Markdown intelligence brief
Delivery mode report-only

Keep the run conservative: prefer recurring-sender detection over one-pass category sampling, use memory to reduce false positives, prefer deduplicated themes over sender-by-sender summaries, and narrow scope rather than expanding into the full inbox when discovery is noisy.

Prompt Inputs

Add context only when you want tighter scope or a different audience, for example:

Newsletter Gmail scope: label:market-newsletters newer_than:7d
Topic priorities: competitors, product launches, tooling, market moves
Audience: founder and product leads
Output emphasis: prefer concrete launches, funding moves, pricing changes, and tooling announcements

Docs