Spamfilter

Spamfilter is a Docker-based Internet Message Access Protocol (IMAP) spam filter for self-hosted mailboxes, scoring each account's mail with a companion…

  • Communication
  • Network Services
  • Productivity
  • Security
  • Tools & Utilities
Spamfilter is a Docker-based Internet Message Access Protocol (IMAP) spam filter for self-hosted mailboxes, scoring each account's mail with a companion Rspamd container, moving spam to Junk, and learning from your own Inbox-to-Junk moves without deleting messages.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

41/100 Fair

Maintained
69last commit 160 days ago
Popular
2110 GitHub stars
Easy to install
351 install platform, install notes, Docker image
Light to run
108 GB minimum RAM

Availability: platform listings confirmed today, links checked 18 days ago

GitHub stars
10
Open issues
0
Last commit
2026-06-14
Latest release
v1.0.0
License
MIT
Activity
stale

Checked 17 days ago - source: GitHub (marcelverdult/imap-spamfilter)

Resources & Compatibility

Minimum requirements: 8 GB RAM, 2 CPU core(s), 100 GB disk.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Install it with Python's package installer Run pip install spamfilter, or use python3 -m pip install spamfilter if your terminal can't find pip. This puts the spamfilter package where Python can import it, so your scripts won't fail on import. source
  • Install the extra you need If you plan to use OpenAI, API, or Transformers features, install the matching extra, such as pip install spamfilter[openai]. Without the right extra, those features may error because their extra dependencies are missing. source
  • Confirm the install worked After installing, run import spamfilter and print spamfilter.__version__ in Python. Seeing a version like 'v3.0.0' confirms the install worked, so you don't have to guess later when your code misbehaves. source
  • Four-container setup on the spamnet network Run spamfilter alongside three companions on a shared Docker network named spamnet: spamfilter-redis (stores rspamd Bayes data), spamfilter-rspamd (scores mail), and spamfilter-unbound (local DNS). Per-account settings (credentials, modes, threshold) go in /mnt/user/appdata/spamfilter/accounts.yml; edit the file by hand and restart the container to apply changes. source
  • Deployment Docker-based; use the included Unraid templates or run on any Linux host with Docker. source
  • Unraid deployment The repository includes Unraid templates for the four-container Docker deployment. source

Fetched today - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Do I need to run more than the main spamfilter container?

Yes, run four containers on a shared Docker network named spamnet: spamfilter-redis, spamfilter-unbound, spamfilter-rspamd, and spamfilter. Redis persists rspamd data, unbound provides local recursive DNS, and rspamd scores messages.

How do I change account modes, thresholds, or credentials?

Per-account settings go in /mnt/user/appdata/spamfilter/accounts.yml. Edit the file by hand and restart the container to apply changes.

Where is the Bayes training data kept?

The spamfilter-redis container stores rspamd Bayes tokens, fuzzy hashes, and neural weights using AOF and RDB. Backing up that Redis data preserves the learned state.

How does it learn spam and ham?

It learns from explicit user moves: Inbox to Junk as spam and Junk to Inbox as ham after the grace window. It can also learn from Junk/Train-Spam and Junk/Train-Ham, and it does not autolearn.

What happens if rspamd is down or a folder is missing?

It fails closed, so the message stays put. It never deletes mail and only uses IMAP MOVE operations.

Answers sourced from github.com

Community

Collected 49 days ago - public community threads