Bobby Network · Website admission
Bring your website into the Bobby Network.
Submit your URL. Bobby will inspect the website, prepare its profile and check whether it is ready to connect.
- No account required
- One URL, no forms
- Private status link
- Evidence with every fact
We ask for information only when we cannot find it ourselves — and submitting a website never guarantees admission.
How it works
Submit, check, discover, review, claim.
Five steps in plain language, each labelled with what actually runs today.
- 01
Submit
Enter a website URL. No account, no form, no category picker.
Live now - 02
Check
The address is normalised and safety-checked, redirects are resolved and the canonical domain is established.
Live now - 03
Discover
Bobby selects up to ten useful public pages by information value and reads what is publicly available.
Partly live - 04
Review
A structured, evidence-backed profile is prepared and checked against Bobby Network standards.
Planned - 05
Claim
If the website looks compatible, an authorised contact may be invited to claim it and verify control.
Planned
Operational state
What runs today, and what is still waiting.
Bobby Website never narrates work it is not doing. Every capability below is listed with its real state.
URL submission and reference
LiveEvery submission is stored durably with a human-readable reference and a private status link.
Safety checks (SSRF, private ranges, redirects)
LiveLoopback, private, link-local, reserved and cloud-metadata targets are refused, and resolved addresses are re-checked across every redirect hop.
Rate limiting and duplicate detection
LiveRepeat submissions are folded onto the same website record instead of starting a second scan.
Canonical domain and robots resolution
LiveRedirects, HTTPS and robots directives are resolved before any page is chosen.
Ten-page selection by information value
LivePages are ranked by role — homepage, about, pricing, contact, terms, privacy, docs, API — not by link order.
Page reading — Firecrawl
Awaiting credentialsThe adapter is built. Without credentials the scan pauses truthfully instead of returning an invented result.
Page screenshots — URLBox
Awaiting credentialsThe adapter is built. Captures will be private by default with source URL and capture time.
Invitations and ownership verification
PlannedThe lifecycle states and records exist; message delivery and verification methods are a later batch.
Public API and MCP server
PlannedThe domain service layer is structured for it. Nothing is published without authentication and governance.
The ten-page scan
Ten useful pages, chosen deliberately — not the first ten links.
The scan resolves redirects and the canonical domain, checks HTTPS, reads robots directives, navigation and the sitemap, then selects the pages that actually carry the information a profile needs. That selection runs on every submission today.
- 01
Homepage
- 02
About
- 03
Product / Services
- 04
Pricing
- 05
Contact
- 06
Brand
- 07
Documentation
- 08
Terms
- 09
Privacy
- 10
API / Authentication
Page reading — Firecrawl
The adapter is built and wired into the scan. Until credentials are added the run parks at provider_configuration_required rather than returning an invented result.
Page screenshots — URLBox
The adapter is built. Captures will be stored privately with their source URL and capture time, and are never shown on a public page.
Website profile
Every material fact carries its source.
A structured profile that separates what was confirmed, what was inferred and what is still unknown.
- Canonical domain
- example-demo-site.test
- Confirmed
- Legal operator
- Not established
- Unknown
- Countries & languages
- EN, NL
- Inferred
- Pricing model
- Subscription
- Inferred
- Terms & privacy
- Both published
- Confirmed
- API
- Documented
- Documented but unverified
- MCP
- —
- Not detected
Fictional demo record. No real website is shown here.
Evidence and screenshots
Each scanned page is recorded with its URL, capture time and confidence. Screenshots stay private by default and are never published.
Show evidence detail for one fact
fact: terms_published · source: /terms · captured: 2026-09-01T09:12Z · confidence: confirmed
Compatibility
An explanation, never a mystery score.
Compatibility is read area by area, in words a non-technical owner can act on.
- Website identityReady
- Legal informationAttention required
- Brand readinessPartial
- Route standardsReady
- APIDocumented but unverified
- MCPNot detected
- AuthenticationDetected but authentication required
- Privacy & data handlingPartial
- Security baselineReady
- Bobby Network connectionsNot detected
“Not detected” means we did not find it on the pages we inspected. It is never proof that a capability does not exist.
Status vocabulary
- Submitted
- Safety check
- Scanning
- Profile generated
- Information incomplete
- Under review
- Changes required
- Eligible to claim
- Invitation sent
- Ownership verification
- Accepted
- Rejected
- Suspended
- Withdrawn
Missing information
We ask only for what we genuinely could not find.
- Contact email — only if no public business address was discovered.
- Legal company name — only if it could not be established.
- Address and country — only when required for admission.
- Company registry number — never required from an individual or unregistered project.
- VAT / tax number — conditional, and skippable where none exists.
Ownership
Eligible to claim is not the same as accepted.
Eligible to claim
The scan suggests the website is compatible and an authorised contact may be invited after a person reviews the discovered contact. One transactional message — no accounts created, no marketing lists.
Accepted
An authorised representative verified control — planned options include DNS TXT, an HTML meta tag, a verification file, domain-email confirmation, Bobby ID authority or controlled manual review — reviewed the profile and completed admission. An invitation alone is never proof of ownership.
Bobby Network
Bobby Website is the system of record for websites.
Website submissions, scans, evidence, canonical profiles, ownership claims and admission status live here. Other products read from it under governance; canonical updates and admission decisions stay with Bobby Website.
Bobby Guru
Would read profiles, scans and evidence through a versioned API and governed MCP server to explain issues and recommend corrections. Not integrated.
Bobby Directory
May publish an accepted website later — only when explicitly authorised.
Bobby Connectors
Governs external provider and credential connections.
Bobby ID
May verify people, companies and authority.
Bobby Dashboard
Reserved for a future combined cross-product dashboard.
planned events · website.submission.created · website.scan.completed · website.profile.ready · website.compatibility.changed · website.changes_requested · website.accepted · website.suspended
Safety & privacy
Controls that run now, separated from commitments that do not.
The first list is enforced on every submission today. The second describes behaviour that arrives with the remaining capabilities.
Enforced today
- Loopback, private-network, link-local, reserved and cloud-metadata targets are refused.
- Resolved addresses are re-checked on every redirect hop to blunt DNS rebinding.
- Only http and https are accepted; every other scheme is rejected outright.
- Submissions are rate limited and folded onto an existing website record when duplicated.
- Robots directives are read before any page is selected.
- A submission is readable only through its own secret status link.
Design commitments, not yet in force
- Screenshots and page evidence kept private by default once providers are connected.
- Discovered contact details used only for one reviewed admission invitation.
- Authority verification required before control or final acceptance.
FAQ
Questions worth asking first.
Submit your URL. We'll discover the rest.
One field is all it takes. No account, no form — and no guaranteed admission.
Back to the URL fieldBobby Website is operated by TorosRicos OÜ.