ChatGPT WordPress diagnostics

Five focused checks

• Current availability: request the public home page and report its response. A single observation does not measure historical uptime or test checkout.
• Security headers: inspect recognised HTTP security settings without returning raw headers or cookies. Their presence does not establish that a website is secure.
• Robots.txt: check basic syntax and crawl rules. Search engines make their own crawling and indexing decisions.
• XML sitemap: validate one public document. Listed pages and child sitemaps are not fetched.
• Plugin version: compare a WordPress plugin slug and exact installed version that you supply with known vulnerability intelligence. Missing matches and unavailable coverage do not establish safety.

Use public information only

Give ChatGPT the public website origin, such as https://example.com, or an exact public XML sitemap address. For a plugin lookup, supply its WordPress.org slug and installed version. Do not guess a version or include passwords, access tokens, signed links or private addresses.

The tools reject credentials, query strings, internal network destinations and unexpected input fields. They cannot sign into WordPress, change a site, read a BugShield account, retrieve Vault credentials or look up another person's diagnostic result.

Diagnostics privacy and data use

BugShield Ltd processes the tool name and public inputs sent by ChatGPT to carry out your requested check. Website checks send ordinary public web requests from our scanner. The destination website receives the scanner's network address and request details, without your browser cookies or WordPress credentials.

Exact plugin lookups use a cached Wordfence Intelligence feed. BugShield retrieves this public advisory dataset using its own provider credentials; the lookup does not send your website address or conversation to Wordfence. Provider notices accompany displayed findings.

Submitted inputs and diagnostic results are processed in memory and are not saved as reports. Results return to the requesting ChatGPT connection. Temporary worker processes end after the check, with a maximum execution time of 55 seconds. BugShield does not retain conversation history through this integration.

Abuse controls store counters with keyed hashes of the connecting client and target domain or plugin slug. These records expire within 24 hours; worker reservations expire within 100 seconds. The public advisory cache expires after one hour. The MCP service disables request-body and access logging. Network infrastructure still processes connection information to deliver traffic.

We use this information to provide diagnostics and prevent abuse, not for advertising or sale. ChatGPT processes and retains your conversation and tool results under your OpenAI settings and its own privacy policy. Removing this integration does not delete an existing ChatGPT conversation.

For data questions, contact [email protected]. Our privacy policy explains BugShield's wider data practices and your rights. This section describes the narrower data use of the diagnostics integration.

Use and limitations

Use the integration for public websites you own or have permission to check. These are bounded, passive observations. Do not use the tools to probe private systems, bypass access controls or submit sensitive information.

Findings support troubleshooting; they are not a security certification, a malware scan or a promise of availability. A failed or incomplete check remains unknown. Review changes with your developer and keep recoverable backups before acting on a result.

Usage limits protect service availability. Checks can be declined when capacity, provider access or safe network access is unavailable. Our terms of use apply alongside these tool-specific limits.

Get help

Contact BugShield with the tool name, time of the check and a description of the problem. Omit passwords, private links and account credentials.