Idea: read-only Lightning node intelligence extension for LNbits

Hi LNbits community,

We’re exploring whether there would be value in a small LNbits extension that gives node operators an independent view of how their node and channels appear from the public Lightning Network.

The idea is deliberately read-only.

An operator could open the extension, identify their public node, and request a BlueHorseShoe assessment. The assessment uses public-network evidence and bounded routing/reachability observations to help answer questions such as:

  • Are my existing channels providing useful onward reach?

  • Are there areas of the network my current peer set struggles to reach?

  • Which existing channels appear useful for reaching different destinations?

  • Before opening another channel, is there evidence that a prospective peer could improve the connectivity I already have?

No seed, admin macaroon, private channel information, wallet balance, or channel-control permission would be provided to BlueHorseShoe.

A possible workflow would be:

LNbits → request BHS assessment → Lightning payment → retrieve report → display results inside LNbits

LNbits would continue doing what it already does well: wallets and Lightning payments. BlueHorseShoe would remain an external, read-only intelligence service.

We already have the underlying report and machine-readable API operating independently of LNbits. A sample report and API contract are available here:

Public sample report:
https://bluehorseshoe.site/BlueHorseShoe/sample-node-report.html

API / operational-summary-v2 contract:
https://bluehorseshoe.site/BlueHorseShoe/nodeanalysisapi.jsp

The current API uses an L402 payment/retrieval flow. For an LNbits integration, we’d also be interested in community feedback on the right pricing model—particularly whether inexpensive individual assessments or refreshes make more sense than a larger one-off report charge.

We’re deliberately asking before building the extension.

Would this be useful to LNbits node operators?

If so, what would be the smallest useful first version you’d suggest we prototype?

Regards,
David
BlueHorseShoe