Docs are not loading

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

When I try to use @DocsDocs I get no documentation at all. Just the “Add Doc” option.

Steps to Reproduce

When I add a message in the chat referring to documentations us@docsng @docs, I have no documentation visible.

Screenshots / Screen Recordings

Operating System

MacOS

Version Information

Version: 2.6.21
VSCode Version: 1.105.1
Commit: fea2f546c979a0a4ad1deab23552a43568807590
Date: 2026-03-21T22:09:10.098Z
Build Type: Stable
Release Track: Default
Electron: 39.8.1
Chromium: 142.0.7444.265
Node.js: 22.22.1
V8: 14.2.231.22-electron.0
OS: Darwin arm64 24.6.0

Additional Information

I also seem to see a spinner when trying to access the marketplace. I have no idea if this is related. It seems to be related to a specific project.

Does this stop you from using Cursor

No - Cursor works, but with this issue

Hi @Nimrod_Oren,

The combination of empty @Docs and a marketplace spinner points to a connectivity issue between your Cursor client and our backend in that specific workspace. A few things to try:

  1. Try a different project to confirm this is workspace-specific

  2. Reload the window in the affected project: Cmd+Shift+P > “Developer: Reload Window”

  3. Check your network: if you’re behind a corporate proxy or VPN, try switching networks temporarily (e.g., mobile hotspot)

Does @Docs work normally in a different project? And does reloading the window fix it?

As I said, I have other projects that work perfectly. I already restarted Cursor (and even my Mac). It doesn’t seem to be a network issue

Since this is isolated to one project and persists after restarting, two things to try:

  1. Test with extensions disabled: open Terminal, cd to the affected project folder, and run cursor --disable-extensions . – if @Docs works now, a workspace-specific extension is the culprit (re-enable one by one to find it)

  2. If still broken, clear workspace state: close Cursor, press Cmd+Shift+G in Finder and go to ~/Library/Application Support/Cursor/User/workspaceStorage/, find the subfolder for the affected project (each has a workspace.json with the project path inside), delete that subfolder, then reopen the project

Both things didn’t work for me

Does this project have a .vscode/settings.json file? If so, could you share its contents? Workspace-level settings (especially any HTTP or proxy configuration) can interfere with backend connectivity in that specific project.

Also, could you open the affected project, go to Help > Toggle Developer Tools, click the Console tab, then clear the console. After that, type @Docs in the chat and share any red errors that appear. That will show us exactly what’s failing.

{
// Python Test Explorer Settings
“python.testing.pytestEnabled”: true,
“python.testing.unittestEnabled”: false,
“python.testing.cwd”: “${workspaceFolder}”,
“python.testing.pytestArgs”: [
“tests”
],
// Python Environment
“python.defaultInterpreterPath”: “${workspaceFolder}/.venv/bin/python”,
“python.terminal.activateEnvironment”: true,
// Terminal Environment Variables - REMOVED to allow .env file to be used
// These were overriding production config when running fastapi dev
// Tests will get their env vars from the autouse fixture in conftest.py
// Python Testing Settings
“python.testing.autoTestDiscoverOnSaveEnabled”: true,
// Editor Settings
“editor.formatOnSave”: true,
“editor.codeActionsOnSave”: {
“source.organizeImports”: “explicit”
},
// Python Formatting
“[python]”: {
“editor.defaultFormatter”: “charliermarsh.ruff”,
“editor.formatOnSave”: true,
“editor.codeActionsOnSave”: {
“source.fixAll”: “explicit”,
“source.organizeImports”: “explicit”
}
},
// Ruff Settings (new format - migrated from legacy ruff-lsp)
// Configuration is in pyproject.toml under [tool.ruff]
“ruff.enable”: true,
“ruff.importStrategy”: “fromEnvironment”,
// Files to exclude
“files.exclude”: {
/pycache": true,
"/*.pyc”: true,
/.pytest_cache": true,
"
/.ruff_cache”: true,
/.mypy_cache": true,
"
/.coverage”: true,
“**/htmlcov”: true
},
// Test Coverage
“coverage-gutters.coverageFileNames”: [
“coverage.xml”,
“lcov.info”,
“cov.xml”
],
“todo-tree.tree.showBadges”: true,
“todo-tree.tree.scanMode”: “workspace only”
}

Your settings are clean — nothing there would cause this.

One more test that should get us to a resolution: open Terminal, cd to the affected project folder, and run:

cursor --user-data-dir /tmp/cursor-test .

This opens the project with a completely fresh Cursor profile (your real settings aren’t affected).

  • If @Docs works in that window: the issue is cached data that wasn’t cleared by the workspace storage reset. I’ll walk you through clearing the right directories.

  • If it still doesn’t work: press Cmd+Shift+I in that test window, click the Console tab, clear it (the :prohibited: icon), then type @Docs in the chat and share a screenshot of any red errors that appear.

I’m not sure what I’m supposed to see. Nothing gets printed

In the window that opened from cursor --user-data-dir /tmp/cursor-test ., when you typed @Docs, did you see documentation listed, or was it still only “Add Doc”?

Please share a screenshot of that screen if you can, so we can get to the bottom of this issue.

It took a lot of time but after a while the new --user-data-dir /tmp/cursor-test Cursor instance showed the docs correctly.

That confirms the issue is stale cached data in your Cursor profile. To fix it in your regular setup, close Cursor completely, then delete these two cache directories:

rm -rf ~/Library/Application\ Support/Cursor/Cache

rm -rf ~/Library/Application\ Support/Cursor/CachedData

Then reopen Cursor and open the affected project. Your settings and extensions will be preserved.

Tried it and for some reason I still don’t see the docs

Let’s do two steps in sequence - go straight to step 2 if step 1 doesn’t fix it, no need to check in between.

Step 1 — Reset the state database:

Close Cursor completely, then run:

mv ~/Library/Application\ Support/Cursor/User/globalStorage/state.vscdb ~/Library/Application\ Support/Cursor/User/globalStorage/state.vscdb.bak

Reopen Cursor and check the affected project. Chat history resets but settings and extensions are preserved. If @Docs loads, you’re done.

Step 2 — Full profile reset (only if step 1 didn’t work):

Close Cursor again, then:

mv ~/Library/Application\ Support/Cursor ~/Library/Application\ Support/Cursor-backup

Reopen Cursor, sign in, and open the affected project. Settings and extensions sync back automatically. If anything is missing afterward, restore from the backup:

cp ~/Library/Application\ Support/Cursor-backup/User/settings.json ~/Library/Application\ Support/Cursor/User/settings.json
cp ~/Library/Application\ Support/Cursor-backup/User/keybindings.json ~/Library/Application\ Support/Cursor/User/keybindings.json

Step 2 fixed the issue however it deleted everything I installed from the Cursor marketplace disappeared and the marketplace itself is in constant loading (never showing)

Anyway, as soon as I added a new document, I lost docs again

Two separate issues to address:

Extensions: Your extensions are in the backup from Step 2. To restore them, run this in Terminal:

cp -R ~/Library/Application\ Support/Cursor-backup/extensions ~/Library/Application\ Support/Cursor/

Restart Cursor and they should be back. For the marketplace not loading, could you try opening a different project and checking if the marketplace loads there? Since the original issue was project-specific, this will tell us if the marketplace problem is too.

Docs disappearing after add: That shouldn’t happen. Could you open the affected project, go to Help > Toggle Developer Tools, click the Console tab, clear it, then add a doc again? If red errors appear in the console right after adding, share a screenshot. That will show us exactly what’s failing when the docs list clears.

Also, could you share the link of what Docs you were trying to add so I could check server-side if there’s anything?

I’m not really concerned about the extenstions. I’ll figure that out. However when I type @docs I see nothing. Not even the docs that are provided with Cursor. I really want to add Databricks documentation | Databricks on AWS

I checked our servers and can see that the Databricks doc indexing attempt just failed on our end — the scraper wasn’t able to process Databricks documentation | Databricks on AWS . That’s a separate issue from the disappearing docs, but it means even once we fix the list, that specific URL may need adjustment (Databricks’ docs site is JavaScript-heavy and can be challenging for our scraper).

For the main issue — @Docs showing nothing after adding a doc — I need to narrow down what’s happening on the client side. Could you try this:

  1. Open Cursor Settings (gear icon) > go to the Docs section under Features/Indexing

  2. Check if any custom docs are listed there

  3. If so, delete all of them, then check if @Docs shows the built-in documentation again

If deleting the custom docs restores the list, that confirms the issue is with how your local doc data interacts with the docs dropdown. If the list is still empty even after deleting custom docs, I’ll need the DevTools console output to see what’s failing:

  • Open the affected project

  • Help > Toggle Developer Tools > click the Console tab

  • Clear the console (click the :prohibited: icon)

  • Type @ in the chat, then type docs

  • Screenshot any red errors that appear in the console

That console output is the one thing that will tell us exactly where the failure is happening, since our server logs show no errors for your account on the docs listing API.

The console looks full of errors. I have a feeling it has something to do with the team marketplace our company put in place. What do you think?

workbench.desktop.main.js:53311 [Extension Host] [CursorMonitor] Policy fetch failed: SyntaxError: Unexpected token ‘<’, ")
at parseJSONFromBytes (node:internal/deps/undici/undici:5852:19)
at successSteps (node:internal/deps/undici/undici:5833:27)
at fullyReadBody (node:internal/deps/undici/undici:4725:9)
at process.processTicksAndRejections (node:internal/process/task_queues:103:5)
at async consumeBody (node:internal/deps/undici/undici:5842:7)
at async fo.fetchPolicy (/Users/nimrodo/.cursor/extensions/elementor.cursor-usage-monitor-0.5.0-universal/dist/extension.js:2324:3711)
dXw @ workbench.desktop.main.js:53311
$logExtensionHostMessage @ workbench.desktop.main.js:53311
_doInvokeHandler @ workbench.desktop.main.js:56169
_invokeHandler @ workbench.desktop.main.js:56169
_receiveRequest @ workbench.desktop.main.js:56169
_receiveOneMessage @ workbench.desktop.main.js:56169
(anonymous) @ workbench.desktop.main.js:56169
_deliver @ workbench.desktop.main.js:55
fire @ workbench.desktop.main.js:55
fire @ workbench.desktop.main.js:53328
(anonymous) @ workbench.desktop.main.js:60819
workbench.desktop.main.js:56229 [transport] Connect error in unary AI connect ConnectError: [internal] certificate signature failure
at t (workbench.desktop.main.js:53347:7841)
at async Object.getEffectiveUserPlugins (workbench.desktop.main.js:43777:4386)
at async Kbv.$getEffectiveUserPlugins (workbench.desktop.main.js:53347:25603) {arch: ‘arm64’, platform: ‘darwin’, channel: ‘stable’, client_version: ‘3.0.12’, layout: ‘unifiedAgent’, …}
error @ workbench.desktop.main.js:56229
t @ workbench.desktop.main.js:53347
await in t
(anonymous) @ workbench.desktop.main.js:53347
GGh @ workbench.desktop.main.js:62
Hwe @ workbench.desktop.main.js:62
unary @ workbench.desktop.main.js:53347
(anonymous) @ workbench.desktop.main.js:43777
$getEffectiveUserPlugins @ workbench.desktop.main.js:53347
await in $getEffectiveUserPlugins
_doInvokeHandler @ workbench.desktop.main.js:56169
_invokeHandler @ workbench.desktop.main.js:56169
_receiveRequest @ workbench.desktop.main.js:56169
_receiveOneMessage @ workbench.desktop.main.js:56169
(anonymous) @ workbench.desktop.main.js:56169
_deliver @ workbench.desktop.main.js:55
fire @ workbench.desktop.main.js:55
fire @ workbench.desktop.main.js:53328
(anonymous) @ workbench.desktop.main.js:60819
workbench.desktop.main.js:56229 [transport] Connect error in unary AI connect ConnectError: [internal] certificate signature failure
at t (workbench.desktop.main.js:53347:7841)
at async Object.getTeams (workbench.desktop.main.js:43777:4386)
at async L4d.getTeams (workbench.desktop.main.js:58017:40087)
at async L4d.refreshMembership (workbench.desktop.main.js:58017:25505)
at async rt (workbench.desktop.main.js:53345:36795) {arch: ‘arm64’, platform: ‘darwin’, channel: ‘stable’, client_version: ‘3.0.12’, layout: ‘unifiedAgent’, …}
error @ workbench.desktop.main.js:56229
t @ workbench.desktop.main.js:53347
await in t
(anonymous) @ workbench.desktop.main.js:53347
GGh @ workbench.desktop.main.js:62
Hwe @ workbench.desktop.main.js:62
unary @ workbench.desktop.main.js:53347
(anonymous) @ workbench.desktop.main.js:43777
getTeams @ workbench.desktop.main.js:58017
await in getTeams
L4d.refreshMembership @ workbench.desktop.main.js:58017
rt @ workbench.desktop.main.js:53345
workbench.desktop.main.js:65 ERR Failed to refresh authentication, scheduling next attempt {error: ConnectError: [internal] certificate signature failure
at t (vscode-file://vscode-app/Applicati…}
error @ workbench.desktop.main.js:65
error @ workbench.desktop.main.js:65
error @ workbench.desktop.main.js:58050
rt @ workbench.desktop.main.js:53345
workbench.desktop.main.js:56229 [transport] Connect error in unary AI connect ConnectError: [internal] certificate signature failure
at t (workbench.desktop.main.js:53347:7841)
at async Object.getTeams (workbench.desktop.main.js:43777:4386)
at async L4d.getTeams (workbench.desktop.main.js:58017:40087)
at async L4d.refreshMembership (workbench.desktop.main.js:58017:25505)
at async rt (workbench.desktop.main.js:53345:36795) {arch: ‘arm64’, platform: ‘darwin’, channel: ‘stable’, client_version: ‘3.0.12’, layout: ‘unifiedAgent’, …}
error @ workbench.desktop.main.js:56229
t @ workbench.desktop.main.js:53347
await in t
(anonymous) @ workbench.desktop.main.js:53347
GGh @ workbench.desktop.main.js:62
Hwe @ workbench.desktop.main.js:62
unary @ workbench.desktop.main.js:53347
(anonymous) @ workbench.desktop.main.js:43777
getTeams @ workbench.desktop.main.js:58017
await in getTeams
L4d.refreshMembership @ workbench.desktop.main.js:58017
rt @ workbench.desktop.main.js:53345
workbench.desktop.main.js:65 ERR Failed to refresh authentication, scheduling next attempt {error: ConnectError: [internal] certificate signature failure
at t (vscode-file://vscode-app/Applicati…}
error @ workbench.desktop.main.js:65
error @ workbench.desktop.main.js:65
error @ workbench.desktop.main.js:58050
rt @ workbench.desktop.main.js:53345
workbench.desktop.main.js:56229 [transport] Connect error in unary AI connect ConnectError: [internal] certificate signature failure
at t (workbench.desktop.main.js:53347:7841)
at async Object.getTeamPrivacyModeForced (workbench.desktop.main.js:43777:4386)
at async workbench.desktop.main.js:58017:35371
at async Promise.allSettled (index 0)
at async produceFn (workbench.desktop.main.js:58017:35331) {arch: ‘arm64’, platform: ‘darwin’, channel: ‘stable’, client_version: ‘3.0.12’, layout: ‘unifiedAgent’, …}
error @ workbench.desktop.main.js:56229
t @ workbench.desktop.main.js:53347
await in t
(anonymous) @ workbench.desktop.main.js:53347
GGh @ workbench.desktop.main.js:62
Hwe @ workbench.desktop.main.js:62
unary @ workbench.desktop.main.js:53347
(anonymous) @ workbench.desktop.main.js:43777
(anonymous) @ workbench.desktop.main.js:58017
produceFn @ workbench.desktop.main.js:58017
await in produceFn
(anonymous) @ workbench.desktop.main.js:132
hc @ workbench.desktop.main.js:132
$ @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:132
Gue @ workbench.desktop.main.js:132
hqe @ workbench.desktop.main.js:132
refetch @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:58017
(anonymous) @ workbench.desktop.main.js:43857
setTimeout
e.setTimeout @ workbench.desktop.main.js:55620
(anonymous) @ workbench.desktop.main.js:43857
shouldHaveGhostModeFromEnterprise @ workbench.desktop.main.js:58017
granularPrivacyModeRawEnum @ workbench.desktop.main.js:58017
isBackgroundComposerEnabled @ workbench.desktop.main.js:58020
onChange @ workbench.desktop.main.js:53203
ef.defer @ workbench.desktop.main.js:462
(anonymous) @ workbench.desktop.main.js:132
hc @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:132
$Zh @ workbench.desktop.main.js:132
E0t @ workbench.desktop.main.js:132
MHn @ workbench.desktop.main.js:132
Mky @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:132
Gue @ workbench.desktop.main.js:132
Lky @ workbench.desktop.main.js:132
Gue @ workbench.desktop.main.js:132
N @ workbench.desktop.main.js:132
P @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:132
Promise.then
$ @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:132
Gue @ workbench.desktop.main.js:132
hqe @ workbench.desktop.main.js:132
refetch @ workbench.desktop.main.js:132
(anonymous) @ workbench.desktop.main.js:58017
(anonymous) @ workbench.desktop.main.js:43857
workbench.desktop.main.js:56229 [transport] Connect error in unary AI connect ConnectError: [internal] certificate signature failure
at t (workbench.desktop.main.js:53347:7841)
at async Object.getTeams (workbench.desktop.main.js:43777:4386)
at async L4d.getTeams (workbench.desktop.main.js:58017:40087)
at async L4d.refreshMembership (workbench.desktop.main.js:58017:25505)
at async rt (workbench.desktop.main.js:53345:36795) {arch: ‘arm64’, platform: ‘darwin’, channel: ‘stable’, client_version: ‘3.0.12’, layout: ‘unifiedAgent’, …}
error @ workbench.desktop.main.js:56229
t @ workbench.desktop.main.js:53347
await in t
(anonymous) @ workbench.desktop.main.js:53347
GGh @ workbench.desktop.main.js:62
Hwe @ workbench.desktop.main.js:62
unary @ workbench.desktop.main.js:53347
(anonymous) @ workbench.desktop.main.js:43777
getTeams @ workbench.desktop.main.js:58017
await in getTeams
L4d.refreshMembership @ workbench.desktop.main.js:58017
rt @ workbench.desktop.main.js:53345
workbench.desktop.main.js:65 ERR Failed to refresh authentication, scheduling next attempt {error: ConnectError: [internal] certificate signature failure
at t (vscode-file://vscode-app/Applicati…)