
- ESX
- QBCore
- vRP
A FiveM resource fails to start when the server cannot find it, cannot read its manifest, starts it before a dependency, hits a Lua error, or denies entitlement. Find the first error naming the resource, then use the matching section below.
Messages are quoted only where a Cfx.re page, a framework or library doc, a seller's doc or a Cfx.re forum post shows them; the source is next to each. Where no exact message could be verified, the section says so. This is a documentation review dated 2026-10-13, not a hands-on test: nothing here was reproduced on a server.
The FiveM script tester has an error explainer that runs in your browser; redact license keys and IPs before sharing any log.
| You see | It points to | Source |
|---|---|---|
Couldn't find resource | Folder or manifest is not where the server looks | Cfx.re forum thread |
does not specify an `fx_version` in fxmanifest.lua then Couldn't start resource ... | Manifest version missing or unsupported | Cfx.re forum thread |
Could not start dependency X for resource Y. | Resource X failed first | Same thread |
No such export ... in resource ... | Provider not running, started later, or export name wrong | ox_inventory docs |
Error parsing script / Error loading script / Failed to load script | Lua syntax or load-time error | Forum, forum |
You lack the required entitlement | Server key and resource owner do not match | Asset Escrow docs |
Native XXXXXXXX is not supported | Outdated server artifact | ESX docs |
Meaning: the Cfx.re scripting introduction says a resource needs a manifest to be detected. No folder with a readable fxmanifest.lua under that name means nothing to start.
Confirm: list the path. You want resources/[optional_group]/resource_name/fxmanifest.lua. Zips can unpack to resource_name/resource_name/fxmanifest.lua. Compare the name after ensure with the folder name, including -, _ and, on Linux, letter case. For a folder added while the server runs, refresh rescans resources and loads manifests (server commands).
Fix: move the inner folder up, correct the name, run refresh, then ensure resource_name. If every custom resource is missing while the default ones work, also check the upload: in the forum thread above, the poster reported that switching FTP client (from WinSCP to FileZilla) fixed it, without knowing why. That is a single anecdote, and the seller docs cited in section 5 report the opposite for protected files (FileZilla corrupting them). Re-uploading the archive and extracting it on the server is the safer test.
Meaning: the manifest reference says a resource must specify a manifest version, and that resources without one currently get a default that is planned to be removed. It lists cerulean, bodacious and adamant as fx_version values, and gta5 and rdr3 among the game values.
Confirm: open fxmanifest.lua and check it has fx_version and game. In the forum thread, replies disagree: one says the bodacious value needs a newer server build (it was added in 2207), another says to download the latest artifact, and a third says they got the same error on every build. Treat an outdated artifact as one possible cause. Also check that each file in client_script, server_script, shared_script and files exists. I could not find a Cfx.re source for the exact message a missing listed file produces, so confirm by comparing the manifest against the folder.
Fix: restore the original manifest from the download, or update the artifact. Writing one: How to Write and Check a FiveM fxmanifest.lua.
Meaning: the manifest's dependency directive requires the named resource to load before the current one. In the forum thread, the log shows the dependency failing with its own warning, followed by Could not start dependency essentialmode for resource esplugin_mysql. The dependent resource is the symptom.
Confirm: scroll up to the first error about the dependency and read it. Then check that the dependency is ensured in server.cfg (conventionally above your resource) and that the manifest's name matches the dependency's folder name.
Fix: repair the dependency; see the other sections for its message. Start-order references: QBCore, ESX, Qbox and Configure FiveM server.cfg and Verify Your Changes.
Meaning: the ox_inventory docs list three likely causes for No such export * in resource ox_inventory: an error occurred while starting ox_inventory or a dependency such as ox_lib; the calling resource starts before ox_inventory; or the export being called does not exist. The same pattern, no such export getPlayerData in resource mumble-voip, appears in a Cfx.re forum thread that has no answer.
Confirm: (1) the named resource started without errors, (2) its ensure line is above the caller's, (3) the export name in the caller matches the provider's code and manifest. The Lua runtime page adds that manifest exports only become available after the first scheduler tick, so treat code that calls one very early as a suspect.
Fix: fix the provider's own error first, then the order, then the name. Never create an empty export to silence the line.
Two shapes appear in the forum threads: Error parsing script ... in resource ...: ... unexpected symbol (the file does not parse) and Error loading script ... in resource ...: ... attempt to index a nil value (global 'ESX') (the file parsed, then failed when it ran), each usually followed by Failed to load script ....
Failed to load script @es_extended/imports.lua came with global 'ESX' being nil. Replies suggested checking how the manifest includes imports.lua and running a newer ESX; the thread records no confirmed fix. Check that the framework starts first and that the script targets the framework version you run.unexpected symbol near '<\239>' in config.lua, then global 'Config' being nil in server/main.lua. The thread gives no explanation. My reading, not a documented cause: \239 is decimal for 0xEF, the first byte of a UTF-8 byte-order mark, and the second error looks like a consequence of the first. Re-save the file as UTF-8 without BOM, then retest.syntax error near <\1> for protected files and pair it with fixes about corrupted or incomplete transfers and outdated servers. Re-upload the original archive, extract it on the server and update the artifact.Fix: go to the file and line in the first message, compare with your backup, and undo your last edit to config or item files. Do not edit protected files.
I could not verify one universal oxmysql failure message, so look for the first database error printed by oxmysql or the script. The oxmysql docs say to set mysql_connection_string with set (“only ever use set”) before starting resources, start oxmysql near the top, and avoid ; , / ? : @ & = + $ # in the connection values because they may be unsupported. ESX's troubleshooting page names “ESX Multicharacter couldn't connect to the database” and attributes it to a wrong connection string in server.cfg or a database connection that is not working.
Confirm: the database server runs, the credentials work in a database tool, and the tables the script needs exist. Details: SQL files and database errors and Migrate mysql-async Resources to oxmysql Safely.
Per the Asset Escrow docs, a resource without entitlement will not start and the console shows You lack the required entitlement. Seller docs add that the resource is tied to the Cfx.re account that owns it, so the server's registration key must come from that account, and a full restart may be needed after the first install. Another seller-documented message is [svadhesive] Failed to verify protected resource ..., linked by the sellers to incomplete restarts, corrupted uploads or a missing .fxap file. Steps: How to Fix the FiveM Required Entitlement Error and What Is a .fxap File in a FiveM Resource?.
I found no documented error message specific to an edition mismatch. The Cfx.re comparison documents differences that can stop a resource working: some Enhanced server files are named differently (for example cfx-server.exe instead of FXServer.exe), resources can ship stream_enhanced beside stream, and Asset Escrow is listed as not implemented yet for Enhanced. If you run Enhanced, confirm that the creator supports the exact release; see FiveM Standard vs Enhanced: Which Version Should You Use?.
The first error line, the resource name and version, framework and version, artifact number, edition, and the relevant server.cfg lines with keys and passwords removed. See startup errors and what support needs.
It is the server saying the resource did not start. The reason is on the lines before it, for example a manifest warning. Read the first error naming the resource.
The resource providing the export did not start, started after the caller, or does not have that export. The ox_inventory docs list these three causes.
A file could not be loaded. The error lines around it name the actual problem, such as a syntax error or a nil framework object. Fix that error, not the generic line.
Yes, where resources depend on each other: start a resource after its framework, libraries and any resource whose exports it calls (the ox_inventory docs list a start order for this reason).





