Repository navigation
💡 New script: chat-channel-digest (Google Chat Spaces weekly activity summary & action item extraction) #572
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Summary
A Google Apps Script that monitors selected Google Chat Spaces, summarizes weekly activity using Gemini, extracts action items and decisions, and delivers a structured digest email — following the proven calendar-to-briefing-doc architecture pattern. Opens an entirely new Workspace surface for the project, covering Google Chat for the first time.
Market Signal
Google Workspace is rapidly elevating Chat as a first-class collaboration surface. The Workspace MCP server (May 2026) adds Chat with search, read, and reply capabilities. Workspace Studio added Chat automation steps in September 2026. No open-source, free, browser-deployable Chat digest tool exists — n8n has Chat nodes but requires self-hosting; Zapier charges per-zap. The competitive landscape (Zapier, Make, IFTTT) treats Chat as just another connector, not a deep integration — leaving room for a purpose-built digest that understands conversational context.
User Signal
The project has 34 existing ideas across Gmail (10+), Calendar (5+), and Drive (4+) — but zero Chat scripts. This is a significant blind spot: Chat Spaces are where small-team decisions, action items, and context live. The calendar-to-briefing-doc script proves strong user appetite for periodic digest emails. Cross-app-weekly-review (#516) acknowledges the multi-app need but does not specifically address Chat content extraction.
Technical Opportunity
The architecture pattern is proven and directly reusable: calendar-to-briefing-doc uses hourly trigger → schedule-in-config → content fetching → Gemini formatting → email delivery. The Chat API (via
UrlFetchAppcalling the Chat REST API, or the emerging ChatApp GAS service) provides Spaces listing, message history, and membership enumeration. The dual-layer GAS/Node pattern with dependency injection supports full testability. The deploy page'sSCRIPT_CATALOGalready handles multi-script configuration — adding a chat-channel-digest entry follows the established extensibility pattern.Assessment
Adversarial Review
Strongest objection: Google Chat may ship its own built-in AI summarization, making a standalone digest script redundant. Chat adoption outside enterprise Workspace plans is limited, narrowing the target audience beyond personal users.
Rebuttal: Built-in Chat summaries are ephemeral and don't extract action items into persistent, archivable formats or integrate with Tasks/Sheets. The script targets small teams (2-10 people) sharing a Workspace plan — the project's natural expansion audience beyond solo users. The gmail-ai-classifier already demonstrates multi-account household deployment, proving the project serves beyond strictly personal use. Even if Google ships native summarization, a structured weekly digest email with extracted action items serves a different purpose than in-app ephemeral summaries.
Suggested Next Step
Prototype a minimal
chat-channel-digestscript: list messages from a single Chat Space for the past 7 days and email a plain-text summary, following thecalendar-to-briefing-docdual-layer architecture (code.gs→src/index.js). Add Gemini-powered summarization and action item extraction as a second iteration.All reactions