Skip to content

Docs: sign a spending grant with the official Multipaz Payment card; the minted card is a test fixture - #470

Merged
TheBlackBit merged 8 commits into
mainfrom
docs/467-official-card-path
Nov 19, 2026
Merged

TheBlackBit merged 8 commits into
mainfrom
docs/467-official-card-path

Conversation

@TheBlackBit

@TheBlackBit TheBlackBit commented Nov 18, 2026 •

Copy link
Copy Markdown
Collaborator

Video: the phone run these docs describe, a grant signed with the official Multipaz Payment card: https://youtu.be/D5lfELSPDgA

Story: Sam builds an AI shopping agent and wants it to buy sneakers while Sam is away. Sam reads the gate README to try signing a spending limit on a phone. Before, it said to mint a custom card with a script, import a file by hand and use a patched wallet. Now it says: install the wallet build, add the Payment card from the official Multipaz issuer, approve.

The only steps a developer follows now (gate README, "Device-signed grants"):

  1. Install the Multipaz Wallet development build: download the debug APK linked in the README (it comes from open pull requests, so there is no release yet), and check its SHA-256 before you install.
  2. Add the SD-JWT "Payment card" from issuer.multipaz.org (SD-JWT is the credential format AP2's spending permission, called delegate, is defined for).
  3. Open the approve link, press "Sign with your wallet", scan the QR code, approve.

Before and after

❌ Before — main (gate README, "Device-signed grants"): "Signing on a real phone needs a wallet that supports AP2's delegate request, and no released wallet does yet … With a patched wallet build … mint one with node tools/demo-pki/mint/mint-dpc-sdjwt.mjs --mpzpass."

✅ After — this change (same section): "Sign a grant on your phone, in three steps: 1. Install a Multipaz Wallet development build … 2. In the wallet, add the Payment card from the official issuer … 3. Open the approve link and press "Sign with your wallet". On a computer, scan the QR code with the phone. Then approve in the wallet." It also names the two same-name cards on the issuer (only the SD-JWT one works) and calls the minted card a test fixture. No code changes.

Which ticket

Fixes #467, part of the epic "Approve a spending grant with the official Multipaz Wallet and its Payment card, with no custom card" (#464). Leaves #413 open (the "no released wallet can sign yet" caveat): the docs now say the same thing, and it stays open until a release can sign.

In plain terms

The docs now tell developers to try phone signing with the card the official Multipaz issuer hands out, not a card they mint themselves. The custom card stays, labelled as a test fixture.

The wallet requirement is not new: main already needed a patched wallet build. This change only swaps the card and names the build.

What you're approving

Text only, in five files. Worst case: a sentence is wrong or out of date. Nothing runs differently. The docs do not claim the card's issuer is verified: a grant stays device-signed, not issuer-verified, and they say the wallet shows the requester as unknown.

How to test

Read the section "Sign a grant on your phone, in three steps" in packages/credentagent-gate/README.md, then follow it on an Android phone with a wallet development build: the grant should end authorized (issue 466 has the recording and the build).


For reviewers — the detail

  • packages/credentagent-gate/README.md: the "Device-signed grants" paragraph about patched wallets is replaced by the three steps, what the phone run showed, and a note that the test card has not been tried on a phone with the current request.
  • docs/guides/testing-on-device.md: a notice at the top sends grant signing to those steps; the .mpzpass import stays for age, loyalty and checkout.
  • tools/demo-pki/mint/README.md: mint-dpc-sdjwt.mjs is labelled a test fixture; the "Unverified" paragraph now says the card has not been matched against the current grant request.
  • examples/quickstart/server.mjs and README.md: the comment and paragraph name the official card and the wallet build.
  • Verified with the root npm run build and npm test (1744 tests).

🤖 Generated with Claude Code

Ever Morales added 2 commits November 17, 2026 21:36
…d card is a test fixture (#467)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
…d note the test card is untested on a phone (#467)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
Comment thread packages/credentagent-gate/README.md Resolved
Comment thread packages/credentagent-gate/README.md Resolved
Comment thread packages/credentagent-gate/README.md Resolved
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: ⛔ Needs changes

This docs-only change tells developers to sign a spending grant on their phone with the official Multipaz Payment card instead of a card they mint themselves. Nothing runs differently, and it never claims the card's issuer is checked. Two sentences in the gate's README are wrong, though, and should be fixed first.

Fix before merging

  1. The new text says the wallet shows the requester as unknown because the server doesn't sign its request. The server does sign it, with a certificate the wallet doesn't trust. A developer who wants their store's name shown would never think to set up a trusted certificate. Reword it using the suggestion on that line.
  2. Two older passages in the same README still say the payment card is one you mint yourself. That now contradicts the new steps, which use the official card. Reword both to say the gate doesn't check who issued the card.

Worth a look

  • Step 3 assumes the approve link is open on a computer. On the phone itself, no QR code appears.
  • The signing page still says a "patched" wallet build works, while the docs now say "development build".
  • The PR description's before/after paraphrases the text instead of quoting it, and uses "SD-JWT" without defining it.
Details for the author
  1. packages/credentagent-gate/README.md:918-919: a README claim the code doesn't back (REVIEW.md §6, docs truthfulness). CONFIRMED. intent-sign/request.ts:75 calls makeReaderCert(origin.rpID, args.readerIdentity), :108 sets client_id: x509_san_dns:<rpID>, and :135-138 signs the request object (ES256, oauth-authz-req+jwt, x5c header). "Unknown" comes from a reader cert the wallet doesn't trust: self-signed when readerIdentity isn't set, and no RICAL (the wallet's list of trusted readers) imported.
  2. packages/credentagent-gate/README.md:631-632 and :888: unchanged lines, made stale by this change (§6, docs truthfulness). CONFIRMED. "self-minted demo credential" conflicts with steps :907-911 and with :922, which calls the minted card a test fixture. The fencing itself is intact: a grant is device-signed, not issuer-verified (intent-sign/verify.ts, request.ts:40). The same wording also sits in code comments (examples/device-signed-grants.mjs:19, intent-sign/request.ts:15, verify.ts:10, types.ts:56), which is lower priority.
  3. Step 3, README.md:912: assumes the cross-device flow. PLAUSIBLE: the issue Record a phone run: approve a spending grant with the official Multipaz Payment card #466 run used a desktop browser, and page.ts:136 names both flows.
  4. intent-sign/page.ts:136: UI copy still says "a patched Multipaz build does". Not in this diff.
  5. Out of scope, noted only: the comment at grants.ts:74 still calls \"page\" today's default, while :532 defaults to \"device\".

Backed by the code: the vct pair urn:emvco:dpc:card:1 / com.emvco.dpc (intent-sign/dcql.ts:28, checked again in verify.ts). The claim_sets choice of credential_id or payment_instrument_id (dcql.ts:45,60, pinned in dcql.test.ts:72). The mint script emits payment_instrument_id and writes dpc.mpzpass. The request is SD-JWT only, so the mdoc card can't match (dcql.ts:48-64). signing: \"page\" exists and devSimulateWalletSignature is exported. The anchor #device-signed-grants--the-wallet-signs-the-grant-first-spec-012 resolves (README.md:839), and the relative links resolve. The hosted-dev-store run ending authorized is confirmed in an issue #466 comment. No Avoid glossary word is used as a trust claim. No security control or completion path changed, so no bypass test is needed. Size is 5 files and about 50 lines. Both commits are DCO-signed.

Coverage: scope full, main..HEAD (b6c96ff, 40a8c0b). Read in full: all 5 changed files' changed sections plus surrounding README sections (lines 620-640 and 839-925); intent-sign/request.ts, dcql.ts, verify.ts, page.ts; grants.ts (signing); mint-dpc-sdjwt.mjs; GLOSSARY.md. Repo-wide grep for stale grant-signing instructions. Dimensions: security invariants and bypass tests (n/a, no code changes), correctness of the doc claims against the code, DX and docs truthfulness, honesty fencing and PR text. Effort: high.

…warn about the mdoc card (review of #470)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
Comment thread packages/credentagent-gate/README.md Resolved
Comment thread packages/credentagent-gate/README.md Resolved
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: ⛔ Needs changes

Since the last review: 1 fixed, 1 still open.

This docs-only change tells developers how to sign a spending grant on their phone with the official Multipaz Payment card. The new commit now explains correctly why the wallet shows the store as unknown. One older contradiction is still there.

Fix before merging

  1. Two older passages in the gate's README still say the payment card is one you mint yourself. The new steps say to use the official card, so the README now contradicts itself. A reader may think the official card is not supported, or that the self-made card is still the main path. Reword both to say only that the gate doesn't check who issued the card. A suggested wording is on the comment.

Worth a look

  • "Seen on a phone" now also covers the repo's own mdoc card. Only the official issuer's mdoc card was actually tried on a phone. The claim itself is still true.
Details for the author

Scope: delta. Commit reviewed: 631054e (new since 40a8c0b). Main changed none of this PR's files.

Fixed: the earlier finding at README.md:918-919 (the old text said the server doesn't sign its request). The new text, now at :921-923, says the request is signed with a self-signed certificate unless readerIdentity is set. CONFIRMED against the code: intent-sign/request.ts:75 calls makeReaderCert(origin.rpID, args.readerIdentity), and intent-sign/routes.ts:106 and grants.ts:371-376 pass the host's readerIdentity through. The anchor #presenting-a-stable-reader-identity-optional matches the heading at README.md:652.

Still open:

  1. README.md:631-632 and :888 still say "self-minted". This is a docs claim the change contradicts (REVIEW.md §6). CONFIRMED. Neither line is in this diff, so the inline comment sits at :923. Fencing is intact: grants stay device-signed, not issuer-verified.
  2. (Worth a look) README.md:913-914: the "seen on a phone" citation, issue Record a phone run: approve a spending grant with the official Multipaz Payment card #466, records only payment_sca_mdoc failing. payment.mpzpass wasn't tried on a phone there. It can't match anyway, because the request is SD-JWT only (intent-sign/dcql.ts:4). CONFIRMED.

Backed by the evidence: multipaz pull request 2196 exists, is open, and builds on 2178, where the shared spending-permission renderer for the wallet's pull request 48 lives. The new testing-on-device.md table row matches the gate README. tools/demo-pki/out/payment.mpzpass exists and is the mdoc org.multipaz.payment.sca.1 fixture (mint/README.md:18). The commit is DCO-signed. No code changed, so no bypass test is needed.

Coverage: read commit 631054e in full, along with README.md:620-675 and :880-930. Checked readerIdentity threading across src/, the issue #466 comments, and multipaz pull request 2196. Dimensions: docs truthfulness and honesty fencing (the only ones a docs-only delta can affect). No sub-agents, since the delta is about 15 lines of docs. Effort: high.

…), step 3 for phone and computer, mdoc citation (review of #470)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: 💬 Ready — small notes

Since the last review: 2 fixed, 0 still open.

This docs-only change tells developers how to sign a spending grant on their phone with the official Multipaz Payment card. The new commit removes the last contradiction: the older passages now say only that the gate doesn't check who issued the card. It is safe to merge.

Worth a look (not blocking)

  • The signing page's own message, shown when the browser can't run the wallet step, still says a "patched" wallet build works. The docs now say "development build". This line isn't part of this change.
Details for the author

Scope: delta. Commit reviewed: 427f19e (new since 631054e). Main changed none of this PR's files.

Fixed:

  1. packages/credentagent-gate/README.md:631-632 and :888 (docs truthfulness, REVIEW.md §6). CONFIRMED. Both now say the gate does not check who issued the payment credential, matching the steps at :909-915 and the test-fixture note at :928. Fencing intact: grants stay device-signed, never issuer-verified (:633-635, :925-926). The only remaining "self-minted" in the README (:855) is a hypothetical ("a self-minted credential passes"), which is still true and not a claim about which card the reader uses.
  2. Earlier worth-a-look on the "seen on a phone" citation, README.md:913-915. CONFIRMED fixed. The citation now covers only payment_sca_mdoc; payment.mpzpass is ruled out by the request being SD-JWT only (intent-sign/dcql.ts), not by a phone run.
  3. Earlier worth-a-look on step 3 assuming a computer, README.md:916-917. CONFIRMED fixed. The QR scan is now conditional on a computer, matching intent-sign/page.ts:8 (the platform mediates same-device vs cross-device).

Still noted (not in this diff): intent-sign/page.ts:136 UI copy says "a patched Multipaz build does".

No code changed, so no bypass test is needed. The commit is DCO-signed.

Coverage: read commit 427f19e in full, along with README.md:620-640 and :850-930; repo README grep for "self-minted"; intent-sign/page.ts flow copy. Dimensions: docs truthfulness and honesty fencing (the only ones a ~10-line docs delta can affect). No sub-agents. Effort: high.

@dzuluaga dzuluaga left a comment •

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the video, Ever! It's great to see the official Payment card signing a grant end to end.

⛔ Needs changes. Two small fixes. The docs now send developers to the official SD-JWT Payment card, and the #466 recording shows a grant ending authorized with it.

Before merge

  1. Merging will close #413. The body says "Does not close #413"; GitHub reads "close #413" as a closing keyword, and the PR now lists #413 as an issue it closes. #413 asks for changes to intent-sign/page.ts and the README's "Delegated purchases" line, which this PR doesn't make. Reword to "Leaves #413 open".
  2. Step 1 can't be followed as written. "Build the wallet from that branch" leaves out the hard part: it needs openwallet-foundation/multipaz#2196 published to mavenLocal first. For now, link an APK of the build you recorded, and say it's a development build of unmerged pull requests. Replacing it with the apps.multipaz.org release is tracked in #472. Also: your #466 build was …-dirty (uncommitted changes). If any of those were needed for signing, put them in openwallet-foundation/multipaz-wallet#48 so the APK matches a branch.

Worth a look

  • One line in the body: the wallet requirement isn't new (main already needed a patched build); this PR changes only the card and names the build.
  • intent-sign/page.ts:136 still says "patched" build. Outside this diff.

@dzuluaga dzuluaga removed their assignment Nov 18, 2026
Signed-off-by: Ever Morales <ever.morales@koombea.com>
Comment thread packages/credentagent-gate/README.md Resolved
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: 💬 Ready — small notes

Since the last review: 0 open from Claude. Both items the maintainer asked for are done.

This docs-only change explains how to sign a spending grant on a phone with the official Multipaz Payment card. The new commit makes step 1 possible to follow: it links a download of the exact wallet build used in the recorded run, and the branch it was built from. It is safe to merge.

Worth a look (not blocking)

  • The wallet download sits on a personal Google Drive, with no checksum or version to check it against. Readers install it on a phone that holds their cards. Add the file's SHA-256 and the version the wallet reports. A project-owned location could also be better.
  • The PR description's step 1 still says only "install a development build". It doesn't mention the download link.
Details for the author

Scope: delta. Commit reviewed: c8d8006 (new since 427f19e). Main changed none of this PR's files.

Maintainer's requests: (1) The "close #413" wording is gone. The body now says "Leaves #413 open", and closingIssuesReferences lists only #467. CONFIRMED. (2) Step 1 now links an APK and says it is a development build of unmerged pull requests. It points to the tracking issue (#472, which exists and is open) for the switch to a release. CONFIRMED. The APK is said to match a branch. TheBlackBit/multipaz-wallet@build/466-recorded-wallet exists, and its head ffc6cfe sits directly on 55db8c2, the commit the recorded -dirty build reports. That fits "an earlier version of #48 plus one fix". It is PLAUSIBLE, not proven, that the build's uncommitted changes were exactly ffc6cfe.

Worth a look:

  1. packages/credentagent-gate/README.md:906: the APK has no integrity pin (no SHA-256, no version string) and is hosted on a personal Drive. CONFIRMED absent. Not a code security control, so it does not block.
  2. The PR body's "only steps a developer follows now" doesn't include the APK link. It is stale against the README. CONFIRMED.

Still noted (not in this diff): the UI copy in intent-sign/page.ts:136 still says "a patched Multipaz build does".

No code changed, so no bypass test is needed. Honesty fencing is intact: grants stay device-signed, and the README still says the gate does not check who issued the card. The commit is DCO-signed.

Coverage: read commit c8d8006 in full, plus README.md:890-934. Read issue #472 and issue #466 (build version), the fork branch via the API, and the PR body and closing references. I grepped the other docs (testing-on-device.md, quickstart README and server.mjs) for stale "build from that branch" wording; none found. Dimensions: docs truthfulness, honesty fencing, PR text. These are the only ones a ~10-line docs delta can affect, so no sub-agents were used. Effort: high.

…self-review of #470)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
Comment thread packages/credentagent-gate/README.md Resolved
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: 💬 Ready — small notes

Since the last review: 1 fixed, 1 still open (not blocking).

This docs-only change explains how to sign a spending grant on a phone with the official Multipaz Payment card. The new commit adds a checksum for the wallet download, a fingerprint readers can compare before installing. I downloaded the file, and its checksum matches the one in the docs. It is safe to merge.

Worth a look (not blocking)

  • The docs say to check the checksum but not how. Name the command for each computer type, and the version number the wallet shows after install, so readers know they have the recorded build.
  • The PR description's step 1 still says only "install a development build" and doesn't mention the download link.
Details for the author

Scope: delta. Commit reviewed: d51e971 (new since c8d8006). Main changed none of this PR's files.

Fixed (earlier worth-a-look): the APK had no integrity pin. packages/credentagent-gate/README.md:908 now gives SHA-256 fc36639981b13b74ececdc156253811f516f365c45dd7467270c7c3e1d9bcc8d. CONFIRMED. I downloaded the linked Drive file (153,827,294 bytes; file reports an Android package) and sha256sum matches. It was hashed only, never run or installed.

Still open (not blocking):

  1. README.md:908: no command given for checking the hash, and no wallet version string. The version is recorded in the issue Record a phone run: approve a spending grant with the official Multipaz Payment card #466 comments: 2026.W46.0-pr-48-176-git-55db8c2-dirty. CONFIRMED absent. Inline comment posted.
  2. PR body, "The only steps a developer follows now", step 1 doesn't link the APK, so it doesn't match the README. CONFIRMED.
  3. Not in this diff, still noted: intent-sign/page.ts:136 UI copy says "a patched Multipaz build does".

No code changed, so no bypass test is needed. Honesty fencing is intact: grants stay device-signed, and the README still says the gate does not check who issued the card. The commit is DCO-signed.

Coverage: read commit d51e971 in full and the README line it changed. Downloaded the APK and checked its hash. Read the PR body and the issue #466 comments for the build version. Dimensions: docs truthfulness, honesty fencing and PR text, the only ones a 3-line docs delta can affect. No sub-agents. Effort: high.

Signed-off-by: Ever Morales <ever.morales@koombea.com>
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: 💬 Ready — small notes

Since the last review: 1 fixed, 2 still open (not blocking).

This docs-only change explains how to sign a spending grant on a phone with the official Multipaz Payment card. The new commit adds the version number the wallet download reports, so readers can confirm they installed the recorded build. That number matches the one in the recorded phone run. It is safe to merge.

Worth a look (not blocking)

  • The docs say to check the download's checksum (a fingerprint of the file) but not how. One line would help, for example sha256sum on Linux, shasum -a 256 on a Mac, certutil -hashfile <file> SHA256 on Windows.
  • The PR description's step 1 still says only "install a development build". It doesn't mention the download link the docs now give.
Details for the author

Scope: delta. Commit reviewed: ad4cc8b (new since d51e971). Main changed none of this PR's files.

Fixed (earlier worth-a-look, half of it): packages/credentagent-gate/README.md:908 now names the version 2026.W46.0-pr-48-176-git-55db8c2-dirty. CONFIRMED it matches both version lines in the issue #466 comments (local run and hosted-dev-store run). It also fits the branch the README links: the previous review found build/466-recorded-wallet sits on 55db8c2. I did not unpack the APK to read its versionName, so that the linked file itself reports this string is PLAUSIBLE (it rests on the README's statement that this is the build that signed in issue #466, and on the SHA-256 the previous review matched).

Still open (not blocking):

  1. README.md:908: no command for checking the hash. CONFIRMED absent.
  2. PR body, "The only steps a developer follows now", step 1 doesn't link the APK, so it doesn't match the README. CONFIRMED (body re-read this run).
  3. Not in this diff, still noted: intent-sign/page.ts:136 UI copy says "a patched Multipaz build does".

No code changed, so no bypass test is needed. Honesty fencing is intact: the README still says grants are device-signed and the gate does not check who issued the card. The commit is DCO-signed.

Coverage: read commit ad4cc8b in full (a 2-line README change). Read the PR body and comments and the issue #466 comments for the recorded version. Dimensions: docs truthfulness, honesty fencing and PR text, the only ones this delta can affect. No sub-agents. Effort: high.

…#470)

Signed-off-by: Ever Morales <ever.morales@koombea.com>
@claude

claude Bot commented Nov 18, 2026

Copy link
Copy Markdown

Claude review

Verdict: 💬 Ready — small notes

Since the last review: 1 fixed, 1 still open (not blocking).

This docs-only change explains how to sign a spending grant on a phone with the official Multipaz Payment card. The new commit names the command that checks the wallet download's fingerprint (its SHA-256 checksum) on Mac, Linux and Windows. All three commands are correct. It is safe to merge.

Worth a look (not blocking)

  • The PR description's step 1 still says only "install a development build". It doesn't mention the download link the docs now give.
Details for the author

Scope: delta. Commit reviewed: a15c933 (new since ad4cc8b). Main changed none of this PR's files.

Fixed (earlier worth-a-look): packages/credentagent-gate/README.md:909-910 now names how to check the hash: shasum -a 256 <file> (macOS), sha256sum <file> (Linux), certutil -hashfile <file> SHA256 (Windows). CONFIRMED each is the standard invocation for that OS and prints a SHA-256 to compare against the pinned fc3663…cc8d. The parenthetical stays balanced and the list item still renders as one step.

Still open (not blocking):

  1. PR body, "The only steps a developer follows now", step 1 doesn't link the APK, so it doesn't match the README. CONFIRMED (body re-read this run).
  2. Not in this diff, still noted: intent-sign/page.ts:136 UI copy says "a patched Multipaz build does".

No code changed, so no bypass test is needed. Honesty fencing is intact: grants stay device-signed, and the README still says the gate does not check who issued the card. The commit is DCO-signed.

Coverage: read commit a15c933 in full (a 1-line README change) and README.md:898-919 around it. Re-read the PR body and comments. Dimensions: docs truthfulness, honesty fencing and PR text, the only ones this delta can affect. No sub-agents. Effort: high.

@TheBlackBit
TheBlackBit requested a review from dzuluaga November 18, 2026 23:40
@TheBlackBit

Copy link
Copy Markdown
Collaborator Author

Both fixed.

  1. The description now says "Leaves Show the "no released wallet can sign yet" caveat where people meet phone signing #413 open", so the merge only closes Docs send developers to a custom payment card instead of the official Multipaz one #467. I also added the line that the wallet requirement is not new: main already needed a patched build, and this PR only swaps the card and names the build.
  2. Step 1 now links the APK of the build I recorded, says it comes from open pull requests, and points to Tech debt: replace the wallet APK link with the Multipaz release on apps.multipaz.org #472 for the release that will replace it. It also gives the version the wallet reports and the APK's SHA-256 with the commands to check it.

About the -dirty: it was one uncommitted change. App.getInstance() ran its initialization cancellable, so the presentment activity restarting its coroutine left the storage half-created and the next attempt failed with KeyExistsStorageException. It is not part of multipaz-wallet#48, and that PR was rewritten after my build, so the APK matches the branch build/466-recorded-wallet in my fork (55db8c2 plus that fix), not the PR head. I can open it as a separate PR on multipaz-wallet if you want it there.

@TheBlackBit TheBlackBit assigned dzuluaga and unassigned TheBlackBit Nov 18, 2026

@dzuluaga dzuluaga left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

✅ Ready. Both blocking items are fixed: the merge now closes only #467, and step 1 links the APK with its SHA-256 and the branch it was built from. Thanks, Ever.

@dzuluaga

Copy link
Copy Markdown
Contributor

A few optional follow-ups, none blocking this merge:

@dzuluaga dzuluaga assigned TheBlackBit and unassigned dzuluaga Nov 19, 2026
@TheBlackBit
TheBlackBit merged commit 51d64d6 into main Nov 19, 2026
11 checks passed
@github-project-automation github-project-automation Bot moved this from In review to Done in CredentAgent Nov 19, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

Docs send developers to a custom payment card instead of the official Multipaz one

2 participants