jevmod

September 21, 2026 · View on GitHub

Moderation toolkit powered by TypeSafe Jev: CLI/SDK/HTTP API/MCP plus optional Discord, Telegram, Reddit, Twitch, and YouTube bots with per-category probabilities and thresholds you own.

At a glance	Details
Source	Source
Maintainer	ohernandezdev. Independently curated; this page is not an upstream submission or endorsement.
Format	Python package `jevmod` 0.2.1 (CLI, FastAPI, bots, optional MCP); product site jevmod.dev.
Requirements	Python ≥ 3.10; `TYPESAFE_API_KEY` (or keyring / `.env`). Platform extras for Discord/Telegram/Reddit/etc.
License	MIT.
Disclosure	AI-assisted catalog review; no affiliation with the maintainer. Listing is not an endorsement. Distinct from catalogued Discord Jev Moderation Bot and Telegram jev_antispam_bot. Offline pytest: 151 passed, 3 failed (missing optional modules in this environment), 32 skipped. No live TypeSafe or connected Discord. Upstream benchmark AUROC figures were not re-run.

When to use

Use it when you want reusable moderation judgments (spam/scam/harassment/nsfw/off-topic/self-harm/doxxing/minors + plain-English rules) across a CLI, API, MCP, or chat bots. Prefer the simpler Discord-only Jev Moderation Bot for a minimal self-hosted bot. Flag-only defaults; destructive actions are opt-in.

How it works

Jev answers category Noul questions; application code owns thresholds and actions (off / flag / delete / timeout). Fails open if Jev is unreachable. Self-harm stays flag-only by design. Bots can tune thresholds from log reactions. MCP and Claude Code skill paths help agents wire jevmod into apps.

Get started

pip install "jevmod[keys]"
jevmod init   # stores a verified key in the OS keyring
jevmod check "FREE NITRO for the first 100!! claim at discord-gifts.ru/nitro"

From source (offline subset):

git clone https://github.com/ohernandezdev/jevmod.git
cd jevmod
git checkout 01063f3e927ddbc3c7de923b73f56dcc33034ca9
pip install -e ".[dev]"
pytest tests/ -q

Live checks send message text to TypeSafe and may incur charges (~vendor-stated $0.04 / 1k messages with all categories—verify current pricing).

Examples and demos

CLI jevmod check sample in README; terminal GIF on upstream docs.
Discord/Telegram/Reddit setup sections; developer HTTP API and MCP.
Offline unit tests for judge/CLI/budget/API paths.

Limits and data handling

Message text goes to TypeSafe on live classification. Bot delete/timeout actions can affect real communities—start with flag-only. Keyring/.env handling is documented; never commit keys. Benchmark comparisons are upstream research—read BENCHMARK.md caveats.

Review and maintenance

Reviewed on 2026-09-21 at commit 01063f3: 0.2.1, MIT. AI-assisted review of README, pyproject.toml, LICENSE, and tests. Offline pytest: 151 passed / 3 failed / 32 skipped in this environment. No live TypeSafe.