Event Guard is a firewall that sits in front of your server events. Most money exploits on FiveM come from a menu firing a server event that trusts whatever the client sends, IE a payout with a made-up amount, a purchase of an item that doesn't exist, or a job event spammed a thousand times.
Add one line to a resource's fxmanifest.lua and every server event it registers is checked before the script sees it: how often each player may call it, and what the arguments must look like (types, number ranges, string lengths, table fields). NaN numbers, giant strings and huge or endlessly nested tables are refused on every protected event without any setup.
If you don't know what an event should take, turn on learning mode for a few hours. Event Guard records what real players send and writes a rule for each event that you paste into the config.
Honeypots catch menus outright. They're fake event names that no script on your server uses, and anyone who fires one is banned. Names that belong to a resource you have installed are skipped, so a real script is never treated as bait.
Every violation adds points to the player's score, and the score fades over time. Cross a step and they're kicked, keep going and they're banned. Staff see all of it in one screen: top offenders, the last violations with a safe preview of what was sent, which rules fire, and the bans.
We tell you plainly what it can't do. FiveM doesn't let one resource stop another one's events, so events from scripts that don't include the library can only be watched and punished afterwards. Also, client side cheats that don't use server events (spawning, god mode) need a client anticheat. The README explains all of it.
What it does
- Protection library: one line in a resource's manifest, no code changes
- Rate limits per event and player, by name or by pattern
- Argument rules: types, ranges, lengths, formats, allowed values, table fields and sizes
- Built-in limits against NaN, infinite numbers, giant strings and huge or deep tables
- Flood guard per resource and across all resources
- Honeypot events, skipped automatically when the real resource is installed
- Escalation by score: kick, ban, and a ban after repeated kicks
- Offenders view with live scores, recent violations and rule hit counts
- Learning mode that writes the rules for you
- Its own ban list checked at connect, or hand bans to your ban script with an export
- Watch mode for scripts you can't edit
- Discord webhook for kicks and bans
- About 1 microsecond added per protected call




