AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Inbox Catchup

skill-mostafa-drz-claude-skills-inbox-catchup · by mostafa-drz

>-

No reviews yet
0 installs
35 views
0.0% view→install

Install

$ agentstack add skill-mostafa-drz-claude-skills-inbox-catchup

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-mostafa-drz-claude-skills-inbox-catchup)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Inbox Catchup? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Inbox Catchup

Scan connected communication sources, build a prioritized briefing, and help triage and respond.

Configuration

Edit these defaults before uploading, or override at runtime when prompted.

sources: all                    # "all" or comma-separated: gmail, calendar, slack
time_window: 4h                 # How far back to scan: 1h, 4h, today, yesterday
priority_contacts: []           # Names/emails that always surface first
reply_tone: professional        # professional, casual, brief
auto_draft: true                # Offer to draft replies for items that need response

Process

1. Detect Available Sources

Check what communication integrations and tools are accessible in this session:

  • Google Workspace — Gmail (unread/recent emails), Google Calendar (upcoming events)
  • MCP servers — Check for any connected messaging tools (Slack, Teams, Discord, etc.)
  • Other integrations — Any other data sources available

Report what was found:

Connected sources:
  ✅ Gmail — via Google Workspace
  ✅ Google Calendar — via Google Workspace
  ✅ Slack — via MCP server
  ❌ Messages — not connected

If no communication sources are detected, explain what integrations are needed and how to connect them (Settings > Connected apps for Google Workspace, or configure MCP servers in Desktop settings). Then stop.

2. Ask Preferences (First Use Each Conversation)

If this is the first time running in this conversation, ask:

> Quick setup for this catchup: > 1. Which sources to scan? (default: all connected) > 2. Time window? (default: last 4 hours) > 3. Any priority contacts to surface first?

Accept brief answers. If the user says "just go" or similar, use the defaults from the Configuration section above.

3. Scan All Sources

For each connected source, gather recent items within the time window:

Gmail:

  • Unread emails in inbox
  • Emails where the user is in To: (not just CC:)
  • Threads with recent replies

Google Calendar:

  • Events in the next 4 hours
  • Events that started in the last hour (may have missed)
  • Events with attachments or agenda docs

Slack (if connected via MCP):

  • Unread DMs
  • Mentions in channels
  • Threads the user is part of with new replies

Other MCP sources:

  • Follow the same pattern: unread, mentions, direct messages

4. Build the Briefing

Organize everything into three priority tiers:

🔴 Needs Reply — Messages directly to the user that appear to expect a response. Sort by sender importance (priority contacts first), then recency.

For each item show:

[Source] From: sender — subject/preview
  Received: time ago
  Context: 1-line summary of what they're asking/saying

🟡 FYI — Worth Reading — Informational messages, CC'd threads, channel activity. No response expected but may be relevant.

For each item show:

[Source] subject/channel — 1-line summary
  Why it matters: brief reason this was flagged

🟢 Upcoming — Calendar events in the next few hours.

For each item show:

[Calendar] event title — time
  With: attendees
  Prep: any linked docs or agenda items

5. Present the Briefing

Show a clean summary:

📬 Inbox Catchup — [date/time]
Scanned: Gmail, Calendar, Slack (last 4 hours)

🔴 Needs Reply (3)
  1. [Gmail] Sarah — "Quick question about the API migration"
     30 min ago — Asking which endpoint to use for the new auth flow
  2. [Slack DM] Alex — "Can you review my PR?"
     1h ago — Wants review on #342 before EOD
  3. [Gmail] Jordan — "Meeting reschedule"
     2h ago — Proposing Thursday instead of Wednesday

🟡 FYI (5)
  • [Slack #engineering] Deploy discussion — team debating rollback strategy
  • [Gmail] Newsletter from TechCrunch — AI funding roundup
  • ... (3 more)

🟢 Upcoming (2)
  • 10:30 AM — Sprint planning (with: team)
  • 1:00 PM — 1:1 with manager (agenda doc linked)

Then ask:

> Want me to help draft replies? I'll take them one at a time, starting with the most urgent.

6. Interactive Triage

For each "Needs Reply" item (in priority order):

  1. Show the full message content
  2. Draft a reply matching the configured tone
  3. Present the draft and ask:
  • Send as-is — send the reply
  • Edit — let the user modify, then confirm
  • Skip — move to the next item
  • Stop — end triage, show remaining items

Never send a reply without explicit user confirmation. Always show the draft first.

7. Wrap Up

After triage (or if the user chose not to reply):

✅ Catchup complete
  Replied: 2 messages
  Skipped: 1 message
  Still pending: 0

  Next event: Sprint planning in 45 min

Guidelines

  • Detect, don't assume — only use sources that are actually connected. Never fabricate messages or pretend a source is available when it isn't.
  • Confirm before sending — every reply must be shown to the user and explicitly approved before sending. No auto-send, ever.
  • Respect priority contacts — if configured, always surface messages from these people first regardless of recency.
  • Be concise in summaries — the briefing should be scannable in under 30 seconds. Save detail for when the user drills into a specific item.
  • Graceful degradation — if a source fails or returns no results, note it and continue with the others. Don't block the entire catchup because one source is down.
  • Privacy-aware — don't summarize or quote message content beyond what's needed for triage. The user may be on a shared screen.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.