COMMUNITY OBSERVEDMetric claims: AUTHOR REPORTEDSOURCE AUDITED🤖 Jev Agent

Warden Personal Email Assistant

A personal assistant that uses Jev to classify email work and trigger proactive analytics.

Warden build screenshots
Warden build screenshots. Original screenshots attached to the author's X post.

Overview & result

Kojo Jeffery shows Warden, a personal assistant for life and homelabbing. A WhatsApp voice note triggers an email-classifier and proactive-analytics workflow using Jev-latest, with Laravel, Paperclip, a daily cron, and edge-tts around the decision layer.

From the original post

Sent a whatsapp voice note to Warden (My Personal Agent, managing life and homelabbing) - to build an email classifier and a simple way to be proactive with my emails with analytics. Classified by @typesafeai Jev-latest, framework should be laravel, cordinate through @papercliping by reporting to the CEO and a daily cron to run the announcement skill using edge-tts. Goodness! all done in 30min - Used Deepseek v4.1 for bootstrapping, now going to send off a task to Astra to make things look beautiful.

— Kojo Jeffery

How Jev fits in the loop

  1. A voice note expresses the assistant task and desired email workflow.
  2. Jev-latest classifies the email-related state into the bounded path needed by Warden.
  3. Laravel, Paperclip, cron, and edge-tts execute the selected workflow and announcement.

How to reproduce

  1. Open the original post and inspect all four attached screenshots as the source evidence.
  2. Separate the assistant's email classification decision from the Laravel, cron, announcement, and text-to-speech execution code.
  3. Treat the reported 30-minute build time as an author claim until a runnable repository is available.

Why this build matters

Jev handles the repeatable classification boundary while the surrounding application owns scheduling, integrations, and speech.

Limitations

  • The post does not provide a public repository or a reproducible benchmark.
  • The exact classification labels and confidence thresholds are not specified in the post.

Patterns