Add experimental silent payment transaction creation support #220

nymius · 2025-09-10T19:39:09Z

Description

This PR adds experimental support for creating silent payment transactions through a new CreateSpTx command. The implementation integrates the bdk_sp crate to enable sending Bitcoin to silent payment addresses.

Key changes:

Adds bdk_sp dependency as an optional feature
Implements CreateSpTx command with support for silent payment recipients
Includes parser for silent payment address:amount pairs
Returns signed transactions ready for broadcast (not PSBTs due to silent payment derivation constraints)

Notes to the reviewers

This feature is marked as EXPERIMENTAL and includes warnings against mainnet use
The command returns signed transactions directly rather than PSBTs because silent payment script pubkey derivation cannot be performed securely in a trustless manner with standard PSBT workflows
RBF is disabled for silent payment transactions (sequence set to MAX)
The implementation handles both single key and extended key signers
Error handling uses temporary .expect() calls that should be addressed in future iterations

Changelog notice

Added: Experimental silent payment transaction creation via CreateSpTx command (feature-gated behind sp flag)

Checklists

All Submissions

I've signed all my commits
I followed the conventional commit guidelines
I ran cargo fmt, clippy and test before committing

New Features

* [ ] I've added tests for the new feature

I've added docs for the new feature

…actions Add experimental silent payment sending capabilities with new CreateSpTx command. This command creates signed transactions directly rather than PSBTs due to current limitations in secure shared derivation. - Add bdk_sp dependency with "sp" feature flag - Implement CreateSpTx subcommand for offline wallet operations - Add silent payment recipient parsing utility - Support mixed recipients (regular addresses + silent payments) - Generate signed transactions ready for broadcast - For the moment is not possible to enable RBF for the created transactions. Note: This is experimental functionality for testing only, not recommended for mainnet use.

tvpeter

Thank you for working on adding this feature @nymius,

I couldn't test because I don't know how to construct sp recipients. I will check the documentation on the library and will come back to test it. In the meantime, I have left some comments.
Thank you

tvpeter · 2025-10-11T12:28:06Z

src/handlers.rs

+            }
+
+            if let Some(utxos) = utxos {
+                tx_builder.add_utxos(&utxos[..]).unwrap();


Can you use the Error enum from the error module to gracefully handle errors rather than unwrapping?

I'm reviewing this unwraps again. I was basing these changes in the CreateTx command, that has the same number of unwraps on the same places.
I can come up with a graceful error handling for this PR, but then we would need to open a new PR to update the other command.
Also, as I'm using the unwraps on the same places, then the Error variants should be as general as possible to reuse them in the CreateTx command, and then this PR wouldn't be doing a single thing, but two.
Any recommendations about how to proceed?

Yeah, I have noticed that the CreateTx command has unwrappings, which we have not updated after adding the Error enum. If you can come up with variants to handle both cases, that would be great. After you have handled the CreateSpTx case, you can add an extra commit to fix the case for CreateTx.

tvpeter · 2025-10-11T12:29:41Z

src/utils.rs

 }

+#[cfg(feature = "sp")]
+pub(crate) fn parse_sp_code_value_pairs(s: &str) -> Result<(SilentPaymentCode, u64), String> {


you can return the Error and use the Generic variant instead of String here too.

Same comment than above, this function is following the style of parse_recipient. The error should be generic in terms of parsing or should be particular to silent payments?

if your library is returning a specific error for this, you can add it to the enum or you can just use the generic variant

tvpeter · 2025-10-11T14:48:30Z

src/commands.rs

+        // Clap Doesn't support complex vector parsing https://github.com/clap-rs/clap/issues/1704.
+        // Address and amount parsing is done at run time in handler function.
+        #[arg(env = "ADDRESS:SAT", long = "to", required = false, value_parser = parse_recipient)]
+        recipients: Option<Vec<(ScriptBuf, u64)>>,


is there a need to add these recipients again since the focus is on sp_recipients?

I want to show that's still possible to send a non silent payment output together with a silent payment output in the same transaction, that's why I left this here.

tvpeter · 2025-10-11T14:49:39Z

src/commands.rs

+        recipients: Option<Vec<(ScriptBuf, u64)>>,
+        /// Parse silent payment recipients
+        #[arg(long = "to-sp", required = true, value_parser = parse_sp_code_value_pairs)]
+        silent_payment_recipients: Vec<(SilentPaymentCode, u64)>,


how can one construct sp_recipients? can you add that to the readme?

Each silent payment recipient is a string of the form <silent-payment-address>:<amount>, for example: sprt1qq0u4yswlkqx36shz7j8mwt335p4el5txc8tt6yny3dqewlw4rwdqkqewtzh728u7mzkne3uf0a35mzqlm0jf4q2kgc5aakq4d04a9l734u5ddn6e:1000 (this is a regtest address, because human readable prefix is sprt).
You cannot get addresses from your bdk-cli because I don't think the receiving side is mature enough yet, but;
SilentPaymentCode is prepared to decode any silent payment encoded address, so I recommend you to take a look in bdk-sp encoding tests to get some fake addresses from there.
If you want to create silent payment addresses dynamically you would have to compile sp-cli2, the cli implemented in bdk-sp, maybe that's too much to add to the README (to ask to compile another tool to get one), what do you think?

What I think is that, if it is possible, you should add deriving SP addresses to this PR. I wanted to test how you implemented it in the CLI v2 but it broke. Else you can add a link to the test addresses so it is easy for users to find them.

But if we add the chances to derive addresses, then we are allowing the creation of transactions locking funds into those addresses. I could restrict this functionality to only derive testnet addresses, so no one lose any real funds.
I mean, is problematic to open the door to receive without implementing the whole functionality to scan for new outputs.

Yes, only testnet addresses are fine. We generally warn users against using this tool for mainnet.

github-project-automation bot added this to BDK-CLI Sep 10, 2025

nymius requested review from notmandatory and tvpeter and removed request for notmandatory September 10, 2025 19:39

nymius force-pushed the feat/sp-psbt-send branch from e0a13af to bc40b76 Compare September 11, 2025 12:09

tvpeter requested changes Oct 11, 2025

View reviewed changes

Uh oh!

Add experimental silent payment transaction creation support #220

Are you sure you want to change the base?

Add experimental silent payment transaction creation support #220

Uh oh!

Conversation

nymius commented Sep 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

tvpeter left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

nymius Oct 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

nymius commented Sep 10, 2025 •

edited

Loading

nymius Oct 18, 2025 •

edited

Loading