Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+22 moreClaude CodeCursorCodex CLIMuseDots
OpenClaw
Manus
Grok
Grok Bot+18 agentsInbox Autopilot
A safe, low-cost e-mail assistant: wakes the model only for new mail, drafts but never sends, trashes but never deletes.
$12
Inbox Autopilot
Example session with this skill installed
Set up my inbox: tell me only about important mail, write an evening digest, and clean up old newsletters.
- Read your context and instructions
- Compiled the inbox autopilot
Change gate (real run on a Gmail account)
$ python3 scripts/inbox_monitor.py | wc -l
331 # sorted unread ids; the model only wakes when this list changes
Jobs (from references/jobs.md)
| Job | Schedule | Model runs |
|---|---|---|
Triage — --monitor-script inbox_monitor.py | every 10 min | only when the unread set changes |
Evening digest — facts from inbox_digest_data.py | 18:00 | once a day |
Cleanup — inbox_cleanup.py --apply (no agent) | 19:00 | never |
Cleanup — always a dry run first
$ python3 scripts/inbox_cleanup.py
query: in:inbox label:Auto/Routine older_than:1d -is:starred -is:important -label:Auto/Attention
matching: 0
dry run — nothing moved (add --apply)
Undo is one command (real run after a large clean-up)
$ python3 scripts/restore_from_trash.py --log trash-log.txt
logged ids: 874, still in Trash: 874
dry run — add --apply to move them back to the inbox
Replies are saved as drafts — nothing is ever sent. Starred and important mail is never touched.
Connects securely to your tools. The creator never sees your data.
About this skill
Your inbox is a source of constant context switching and technical debt. This skill implements a production-ready, three-tier assistant that triages incoming mail, generates daily digests, and performs verifiable cleanups using a low-cost, script-gated architecture.
What it does
- Deterministic Triage uses a script-based gate to wake the model only when new mail arrives, preventing expensive empty runs.
- Smart Classification categorizes mail into routine (newsletters, receipts) or attention-required (clients, deadlines) categories.
- Automated Drafting prepares replies for important messages without sending them, keeping you in control.
- Evening Summarization groups a day of JSON-extracted mail facts into a security-first categorized digest with clear action items.
- Verifiable Cleanup moves older routine mail to Trash while logging every message id so a single command can restore it.
How it works
- Setup runs the credential script to securely store IMAP environment variables and test the connection.
- Monitor executes a read-only script to detect new unread UIDs before invoking the LLM.
- Process triggers the agent to label messages, mark them as read, and prepare summaries or drafts.
- Digest & Purge runs an evening cycle to summarize activity and trash routine items older than 24 hours.
Frameworks & tools
Built with Python and Shell scripts using standard libraries. Compatible with Hermes Agent, cron, or any task scheduler. The change monitor and digest work with any IMAP inbox (verified TLS); cleanup and restore use Gmail search and are Gmail-only.
Why this beats prompting it yourself
Prompting alone leads to high token costs from constant polling and "no new mail" responses. This skill provides a deterministic gate and a logged, reversible cleanup mechanism that ensures no important mail is permanently lost to aggressive AI filtering.
Use cases
- Manage a high-volume professional inbox without missing critical client offers.
- Consolidate noisy newsletters and platform notifications into a single evening report.
- Keep the inbox short by moving verified routine mail to Trash, recoverable for 30 days.
- Monitor for security-critical mail like password resets or new login alerts.
Known limitations
Requires an app-specific password or OAuth setup. IMAP UIDs must be mapped correctly to Gmail API IDs. The cleanup script only targets messages older than 24 hours to prevent digest gaps.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scanned
Verified clean 3 days ago
- Passed all security checks, Safe to install
Requires
Python