I have disabled “Include PRs in Search” and that seems to have speed up the indexing again…
Hey folks, we raised an internal incident to address this, it should be slowly starting to improve. Indexing is done in the backend, so this should be version independent as some folks have pointed out!
To put things in a good state, you may want to restart cursor + restart the sync and check if that speeds things up here. Seems to be working well on my side now.
I haven’t noticed any improvement in the index speed.
I deleted my existing index, restarted cursor and rebuild the index. It finished in 5 minutes.
After deleting my existing index and attempting to rebuild it, I immediately encountered a ‘Handshake failed’ error. In the logs, I noticed 'STATUS_FAILURE', 'codebases': [].
2025-07-08 22:32:46.770 [info] Handshake timing: 265.24762499999997, response: {"status":"STATUS_FAILURE","codebases":[]}
2025-07-08 22:32:50.777 [warning] Retrying handshake with timeout 8000. Error: FastRepoInitHandshakeResponse_Status.FAILURE in handshakeWithRetry
2025-07-08 22:32:50.777 [info] Handshake start
2025-07-08 22:32:51.377 [info] Handshake timing: 602.8627090000009, response: {"status":"STATUS_FAILURE","codebases":[]}
2025-07-08 22:32:59.384 [warning] Retrying handshake with timeout 16000. Error: FastRepoInitHandshakeResponse_Status.FAILURE in handshakeWithRetry
2025-07-08 22:32:59.385 [info] Handshake start
2025-07-08 22:32:59.974 [info] Handshake timing: 590.500333, response: {"status":"STATUS_FAILURE","codebases":[]}
Same here
I’m also experiencing the “Handshake failed” issue on indexing. Started this morning.
Same here, but for indexing my docs. I don’t think that all of them have only 10 pages for indexing, and some show very old dates that don’t correspond to when I actually indexed them:
I tried again, and some are do working, but others are failing or just indexing a single page. I’d say the ration of success is 50/50:
Error still persists. Been experiencing this for over the last 6 hours
I also tried switching between different HTTP Compatibility Modes
.
I understand that Cursor Indexing & Retrieval
is a backend service but I’m also including my About Cursor
:
Version: 1.2.2 (Universal)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:07:27.002Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.5.0
Yep. Same here. It seemed to be working, but now it shows “Handshake failed”.
P.S. status.cursor.com is useless
Describe the Bug
Like the title says, I got an indexing problem. I was using Cursor the way I always did until I opened a new folder. After that Cursor started setting up indexing and I noticed it was talking longer than usual. It seems like the indexing never starts until eventually it fails (check screenshot). I am not sure if it’s an issue due to the latest update, this issue appeared after updating Cursor today to the below version:
Version: 1.2.2 (Universal)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:07:27.002Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.4.0
Any ideas as to why this is happening?
Steps to Reproduce
Cursor>Settings>Cursor Settings>Indexing & Docs
Expected Behavior
The indexing should start and my codebase should be synchronized.
Screenshots / Screen Recordings
Operating System
MacOS
Current Cursor Version (Menu → About Cursor → Copy)
Version: 1.2.2 (Universal)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:07:27.002Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.4.0
Does this stop you from using Cursor
Yes - Cursor is unusable
Got the same bug (my post was still pending approval):
Describe the Bug
When opening a newly cloned repository in Cursor for the first time, the codebase indexing system fails consistently with handshake timeouts to Cursor’s indexing servers. The servers respond with immediate STATUS_FAILURE
messages, indicating active rejection rather than connectivity issues. Git indexing completes successfully, but codebase indexing fails completely, preventing code intelligence features from working properly. This appears to be a server-side issue as extensive local troubleshooting has not resolved the problem.
Steps to Reproduce
- Clone a new repository locally to
C:\Users\[user]\[project]
- Open the project in Cursor for the first time
- Observe indexing process in the logs
- Notice that Git indexing succeeds but codebase indexing fails repeatedly
- Check logs showing handshake failures with increasing timeout intervals (4s → 8s → 16s → 32s → 64s → 128s)
- All handshake attempts return
{"status":"STATUS_FAILURE","codebases":[]}
Expected Behavior
- Cursor should successfully establish connection with indexing servers (
https://repo42.cursor.sh
andhttps://api2.cursor.sh
) - Codebase indexing should complete successfully for new repositories
- Code intelligence features should be available immediately
- Both Git indexing and codebase indexing should work without errors
Operating System
Windows_NT x64 10.0.19045
Current Cursor Version (Menu → About Cursor → Copy)
Version: 1.2.2 (user setup)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:19:13.016Z (1 day ago)
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.19045
Additional Information
Repository Details:
- Project Type: React/TypeScript with Vite, Tailwind, and Supabase
- Git Status: Properly configured (
git remote -v
andgit config --get remote.origin.url
work correctly) - Repository is newly cloned, first time being indexed by Cursor
Error Patterns:
- Consistent immediate rejections:
{"status":"STATUS_FAILURE","codebases":[]}
- Quick response times (150-800ms) indicate good connectivity but active server rejection
- Final timeout:
Error: timeout in handshake with retry
- Git submodules error:
ENOENT: no such file or directory, open '.gitmodules'
(expected, as project has no submodules)
Key Technical Observations:
- Git indexing works perfectly: “Secondary handshake successful. Found 1 histories” and “Job completed. Total indexed: 0 PRs + 37 ignored + 0 failed”
- Codebase indexing fails immediately with server rejections (not connectivity timeouts)
- Hash computation succeeds:
1a8ee40504ebaf9f8f9a8233375e32e1067875170466f2c747ebf4f70dbffbd3
- Path key hash consistent:
0cfc51acd171865f03ab669a4aebe99586e0973963972c7910c9b671050948b3
Troubleshooting Steps Attempted:
Restarted Cursor completely - Issue persists
Cleared Cursor cache - Deleted
%APPDATA%\Cursor\User\workspaceStorage\
and%APPDATA%\Cursor\CachedExtensions\
- Issue persistsTested different networks - Tried multiple networks with no firewall restrictions - Issue persists
Verified Cursor sign-in status - Confirmed signed in properly - Issue persists
Disabled/re-enabled indexing - Toggled
cursor.general.enableIndexing
setting - Issue persistsVerified Cursor Pro/subscription status - Account status confirmed - Issue persists
Sample Error Logs:
2025-07-08 18:31:11.051 [info] Handshake timing: 806.8173999999999, response: {"status":"STATUS_FAILURE","codebases":[]}
2025-07-08 18:31:13.436 [info] Handshake timing: 348.5629999999983, response: {"status":"STATUS_FAILURE","codebases":[]}
2025-07-08 18:33:18.634 [error] Error: timeout in handshake with retry
2025-07-08 18:33:18.637 [error] Handshake failed.
Evidence This is Server-Side Issue:
- Immediate
STATUS_FAILURE
responses (not connectivity timeouts) - Quick response times (150-800ms) indicate good network connectivity
- Multiple network configurations tested with same result
- Git indexing succeeds while codebase indexing fails
- Extensive local troubleshooting shows no local configuration issues
- Problem started around 2025-07-08 18:04 and persists consistently
Does this stop you from using Cursor
Yes - Cursor is unusable
The same on Linux
Unable to index the project also.
Describe the Bug
When opening a newly cloned repository in Cursor for the first time, the codebase indexing system fails consistently with handshake timeouts to Cursor’s indexing servers. Git indexing completes successfully, but codebase indexing fails with STATUS_FAILURE
responses and eventual timeout errors. This prevents code intelligence features from working properly.
Steps to Reproduce
- Clone a new repository locally to
C:\Users\[user]\[project]
- Open the project in Cursor for the first time
- Observe indexing process in the logs
- Notice that Git indexing succeeds but codebase indexing fails repeatedly
- Check logs showing handshake failures with increasing timeout intervals (4s → 8s → 16s → 32s → 64s → 128s)
Expected Behavior
- Cursor should successfully establish connection with indexing servers (
https://repo42.cursor.sh
andhttps://api2.cursor.sh
) - Codebase indexing should complete successfully for new repositories
- Code intelligence features should be available immediately
- Both Git indexing and codebase indexing should work without errors
Operating System
Windows 10/11
Current Cursor Version (Menu → About Cursor → Copy)
Version: 1.2.2 (user setup)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:19:13.016Z (1 day ago)
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Windows_NT x64 10.0.19045
Additional Information
Repository Details:
- Project Type: React/TypeScript with Vite, Tailwind, and Supabase
- Git Status: Properly configured (
git remote -v
andgit config --get remote.origin.url
work correctly) - Repository is newly cloned, first time being indexed by Cursor
Error Patterns:
- Initial error:
NoWorkspaceUriError
andGit command failed: git config --get remote.origin.url
(despite git working correctly) - Handshake failures: Consistent
{"status":"STATUS_FAILURE","codebases":[]}
responses - Final timeout:
Error: timeout in handshake with retry
Key Observations:
- Git indexing works perfectly: “Secondary handshake successful. Found 1 histories” and “Job completed. Total indexed: 0 PRs + 37 ignored + 0 failed”
- Codebase indexing fails completely with server connection issues
- Issue is reproducible across multiple restart attempts
- User has opted out of PR indexing via configuration
Sample Error Logs:
2025-07-08 18:06:09.676 [info] Handshake timing: 1699.5946000000004, response: {"status":"STATUS_FAILURE","codebases":[]}
2025-07-08 18:11:19.661 [error] Error: timeout in handshake with retry
2025-07-08 18:11:19.662 [error] Handshake failed.
Does this stop you from using Cursor
Yes - Cursor is unusable
Describe the Bug
Not sure why, but my indexing has been taking hours for some reason. It’s only at 41%. It’s causing major issues in the quality of output, as well as the speed of output, while in agent mode.
Steps to Reproduce
Just try indexing a codebase. It’s happening on all of my codebases.
Expected Behavior
It used to index my codebase in seconds, now it’s taking hours. The bigger the codebase, the longer it takes.
Screenshots / Screen Recordings
Operating System
MacOS
Current Cursor Version (Menu → About Cursor → Copy)
Version: 1.2.2
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:08:52.104Z
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.6.0
Does this stop you from using Cursor
Sometimes - I can sometimes use Cursor
Same…
Codebase Indexing Failing. After along night of cursor continuously lying to me about edits it made, I update to what appears to be the same exact version (1.1.6). Now indexing is failing. I go back to the “older version” and the same thing happens. Upgraded to 1.2.2, same thing happening.
failing for me as well
Same issue for me too.
This is my cursor version: Version: 1.2.2 (Universal)
VSCode Version: 1.99.3
Commit: faa03b17cce93e8a80b7d62d57f5eda6bb6ab9f0
Date: 2025-07-07T06:07:27.002Z (1 day ago)
Electron: 34.5.1
Chromium: 132.0.6834.210
Node.js: 20.19.0
V8: 13.2.152.41-electron.0
OS: Darwin arm64 24.5.0
Issue started happening right after the update.