Signal

Privacy

Last updated 8 October 2026

This privacy notice applies to the Signal form and its MCP tools. They do not cover visiting the rest of the SR3H website or contacting SR3H by email. For questions about those activities, contact hello@sr3h.uk.

Signal processes the public website URL and any optional business context you submit so it can inspect website access and clarity. The SR3H website form can also run a small branded and unbranded discovery sample. The Signal plugin can also run and save independent GPT web-search studies using SR3H’s configured API allowance.

What the tool accesses

It requests the supplied public webpage, up to four relevant pages on the same website, and may request that website’s robots.txt, sitemap.xml and llms.txt files. The website operator can see those requests, including the checker’s user agent and network information.

AI-assisted analysis

When the live sample in the SR3H website form is available, bounded public-page extracts and the optional context you supplied are sent to OpenAI to assess six relevant questions and run a small web-search sample. The request is sent with API response storage disabled. Saved Signal studies use that API access for fresh searches, separate assessment of captured answers and evidence-informed analysis. Neutral search requests contain the question and fixed settings; no personal chat history or business briefing is supplied. The target business is supplied separately after the raw answer is captured. The earlier within-chat sample remains available and uses the host's research tools. Do not submit personal, private or confidential information.

Storage and logs

The tool has no user accounts. Daily usage controls store counters, checked website hostnames and a visitor identifier. When configured, this identifier is derived from your IP address using a secret key and changes each UTC day; otherwise requests share a common limit. Raw IP addresses, full submitted URLs, business context and fetched page contents are not written to this quota store. Its previous daily record is replaced when the next day's check arrives.

A completed website discovery result, including its questions and findings, may be held in Cloudflare’s edge cache for up to 24 hours to avoid repeating the same paid check. Application error messages use general error codes. Cloudflare and OpenAI may process network and operational data under their own terms. Disabling API response storage does not remove their separate operational processing.

Saved visibility studies

A saved study stores the confirmed public business context, checked website evidence, frozen questions and settings, dated raw answers and sources, assessments, suggested interventions, chosen change records and before-and-after results. These records are retained until you delete the study using Signal’s delete tool. Each study is accessed through a randomly generated private reference. Anyone with that reference can access its records; keep it in your private chat and out of public reports. There is no public study list or user-account login for these records. Cancelling a run retains completed answers; deleting a study removes its records and stops its runner.

SR3H does not sell submitted information or use it for advertising or user profiling.

Use public information only

Do not submit private URLs, credentials, personal data or confidential information. Local, private-network and credential-bearing URLs are rejected.

Contact

For privacy questions, email hello@sr3h.uk.