Mail Sniffer Agent

Mail Sniffer is a Python command-line agent concept for organizing job-search email. It is designed to classify incoming messages, extract interview details, update a Notion tracker, create calendar events, and prepare reply drafts for human review.

Recruiter messages, interview invitations, scheduling changes, and application updates arrive across long email threads. Manually transferring details into a tracker and calendar adds overhead and increases the chance of missing a follow-up.

Receive: Gmail push notifications through Google Cloud Pub/Sub are intended to trigger processing when new mail arrives, rather than relying on continuous inbox polling.

Classify: categorize messages such as interview invitations, scheduling, status updates, recruiter contact, paperwork, or spam.

Extract: identify structured details, including interview dates and times.

Update: create or update job-tracker cards in Notion and prepare relevant Google Calendar events and preparation blocks.

Draft for review: generate suggested replies and place them in a review queue. Messages are not intended to be sent without explicit human approval.

Brief: compile a daily summary of job-search activity.

The repository contains approximately 940 lines of Python organized into agent and integration modules for Gmail, Google Calendar, Notion, Pub/Sub, and an OpenRouter-configured LLM client. Configuration is handled through environment variables, with a guided setup script.

This is a prototype design; live Gmail, Calendar, Notion, and Pub/Sub integrations have not been tested against real accounts.

A Google Cloud project, OAuth credentials, and API keys are required to run the integrations.

There is no automated test suite.

Classification and extraction accuracy have not been measured.