Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
When an AI coding assistant fails, first identify whether the problem is your account, the editor or CLI integration, the network, the project, or a usage limit. That keeps you from reinstalling software when the real issue is a missing entitlement, an expired sign-in token, a blocked connection, or a feature setting.
First, identify what is failing
Record the exact error message and note whether the failure affects inline suggestions, chat, or agent actions. Also establish its scope: one file or language, one workspace, or every project. Check the vendor’s service-status page before changing local settings; GitHub specifically recommends checking GitHub status when Copilot is slow.
These clues help narrow the likely cause:
- One feature is missing: Check whether it is enabled, included in your account, supported by your client version, or restricted by an organization policy.
- One file or language is affected: Check per-language settings, file exclusions, and whether the same feature works in another file.
- Only one workspace is affected: Try a small file or a clean workspace. Project size, dependencies, and configuration may be involved.
- Every project is affected: Check sign-in, entitlement, service status, connectivity, and the installed client or extension version.
- The tool is slow rather than unavailable: Compare a small project with the affected one, check local CPU and memory use, and inspect the connection and logs.
For Copilot-specific guidance, see GitHub’s common-issues guide and slow-response troubleshooting guide.
Check account access and usage limits
Confirm that the IDE, extension, or CLI is signed in to the account—and, where relevant, the organization—that has access to the product. A connection or activation message does not necessarily mean the vendor is down: for Copilot, GitHub says it can mean there is no Copilot plan or the client could not obtain a token from the GitHub API. GitHub recommends signing out and back in to request another token.
#1 Best Overall
In VS Code, check the Copilot dashboard, the active account, the feature setting, and the account’s allowance. A temporary rate limit is not the same as exhausting a monthly allowance: the VS Code FAQ says a request can be rate-limited even when monthly AI credits remain, and recommends waiting briefly before retrying. Follow the exact message and your account’s current usage details; reset timing and allowances vary by product and plan. See the VS Code Copilot FAQ.
Update the editor and integration
A feature can disappear or stop working when the editor, extension, or plugin is out of date—or when a current extension no longer supports an older editor. Update the editor and its AI integration, then restart and check the vendor’s current compatibility guidance. GitHub says current Copilot Chat versions align with the latest VS Code release, so an older VS Code installation may not run the latest Copilot Chat. Feature availability can also depend on prerequisites or an organization policy.
Claude Code installation matters
Anthropic describes Claude Code as a terminal agent that can read a repository, edit files, and run commands; IDE extensions are also available for VS Code and JetBrains. Its update route depends on how it was installed: native installations update in the background, while Homebrew and WinGet installations require upgrades through those package managers. Check the Claude Code user FAQ for current details.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Codex configuration errors can be version-specific
For a Codex configuration or version error, first identify the surface you are using and its installed version. OpenAI documents that specified newer versions no longer support approval_policy = "untrusted". For that specific error, its help page says to remove the setting, check profiles and managed configuration for other references, and restart after changing the configuration. Do not apply this fix to unrelated errors. See OpenAI’s Codex help page.
Rank #3
Cursor troubleshooting
Cursor’s troubleshooting link points to general documentation rather than a verified step-by-step guide. Start with its current troubleshooting documentation for product-specific instructions; avoid assuming that another editor’s fixes apply unchanged.
Isolate network, extension, and workspace problems
GitHub’s Copilot slowness guidance identifies network conditions, local performance, editor configuration, proxies, and firewalls as possible causes. The following checks are useful diagnostic hypotheses for other networked coding tools too, but are not vendor-guaranteed fixes for every product.
- Test the connection: Check whether the internet connection is stable. Consider whether a VPN, proxy, firewall, or security product that inspects traffic could be blocking or delaying requests. If possible, compare with another network or device.
- Isolate editor conflicts: Temporarily disable other extensions, especially AI assistants, linters, formatters, and code-analysis plugins. If the problem stops, re-enable them individually to identify a conflict.
- Compare project size: Try a small file and a new minimal workspace. If those work, investigate the original project’s size, dependencies, and configuration.
- Check local resources: If the editor itself is sluggish, inspect CPU and memory use while reproducing the issue.
- Review logs: Look for errors that coincide with the failed request before making more changes.
For Copilot, GitHub’s slowness guide includes product-specific troubleshooting advice.
Free tools Windows power users keep installed
One-click scans. No signup required.
Find diagnostic logs and prepare a useful support report
Capture evidence while the problem is happening. In VS Code, GitHub directs users to the Output panel and the GitHub Copilot channel; in JetBrains, inspect the IDE logs. For other tools, use the vendor’s current logging instructions rather than assuming the same panel or log location exists.
Include the information relevant to your case:
- The exact error text and the feature that failed.
- Editor, extension, plugin, or CLI versions, plus the installation method where it affects updates.
- Steps to reproduce the problem and whether it happens in a clean workspace or only in one project.
- Relevant log entries and, when useful, an example file or language.
- For Codex support cases, OpenAI asks for the exact error, request ID, timestamp, and time zone when available.
Redact passwords, authentication tokens, private code, and personal information before sharing logs or diagnostic files. GitHub’s Copilot slowness guide covers log locations and information to provide; Codex support details are in OpenAI’s help page.
What bug studies can—and cannot—tell you
A 2026 study by Ruixin Zhang, Wuyang Dai, Hung Viet Pham, Gias Uddin, Jinqiu Yang, and Song Wang examined more than 3,800 publicly reported bugs in the Claude Code, Codex CLI, and Gemini CLI repositories. The authors classified more than 67% as functionality-related and 36.9% as stemming from API, integration, or configuration errors. Those figures describe the paper’s selected public bug reports and its classifications; they do not measure how often users encounter problems, compare the products’ reliability, or establish the prevalence of any issue across all users. Read the study on arXiv.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
