[maxpetretta]

Note

This PR was written by Claude Opus 5.5 (Anthropic's model, running in Claude Code) on behalf of @maxpetretta, who asked for the fix and confirmed it works on their Mac.

Description

Since the 2026-05-27 update (#28183, "Fix Time Sensitive iMessage codes"), the iMessage source finds no codes when message.text is NULL and the body only exists in attributedBody. On macOS 26 that is most incoming messages: 388 of the 400 most recent on the test machine.

The query read the body with cast(message.attributedBody as text). SQLite cuts that cast off at the first NUL byte, and in the typedstream blob that byte comes just before the message body. Every such message reached the extension as streamtyped…NSMutableAttributedString, extractCode found nothing, and the list showed the empty view.

Changes:

• Fetch attributedBody with hex() and decode the NSString payload from the typedstream in decodeAttributedBody (src/utils.ts). Handles 1, 2 and 4 byte lengths.
• combineMessageText keeps the Fix iMessage Time Sensitive code detection #28183 behaviour: a notification label such as "Time Sensitive" in message.text is still combined with the decoded body.
• The code-candidate check and search now run on the decoded text in useMessages, since SQL can no longer see the body. The candidate check is the same as the old globs, which all reduce to three consecutive digits or capitals.
• The query scans up to the 1000 most recent messages in the look back window, and the 100 result limit is applied after filtering, so newer unrelated messages can't hide a code. One poll at the full 1000 rows takes about 13 ms.
• New tests build real typedstream blobs and go through the hex path. The existing binary-data tests use hand-written strings that never pass through SQLite, which is how this regression got past them.

Fixes #29673. Probably also the cause of #28593. Related: #21910.

Testing

• macOS 26.7.1, Raycast 2.10.1: the store version showed an empty list. With this branch running through npm run dev, a newly received code appeared.
• Against the same machine's chat.db: the old query path found a code in 0 of the last 400 incoming messages, and the decoded text found one in 31. All 24,889 attributedBody blobs from the past year have the NSString layout the decoder expects.
• npm test (47 passed), npm run lint (0 errors, existing warnings only), npm run build.

Screencast

No UI changes. The list shows codes again.

Checklist

• I read the extension guidelines
• I read the documentation about publishing
• I ran npm run build and tested this distribution build in Raycast
• I checked that files in the assets folder are used by the extension itself (no asset changes)
• I checked that assets used in the README are located outside the metadata folder if they were not generated with our metadata tool (no README changes)

🤖 Generated with Claude Code

[raycastbot]

Thank you for your first contribution! 🎉

🔔 @yuercl @thomaspaulmann @arronhunt @hitolaus @taschaub @omaerkhan @clins1994 you might want to have a look.

You can use this guide to learn how to check out the Pull Request locally in order to test it.

📋 Quick checkout commands

BRANCH="fix/imessage-2fa-tahoe-blank"
FORK_URL="https://github.com/maxpetretta/extensions.git"
EXTENSION_NAME="imessage-2fa"
REPO_NAME="extensions"

git clone -n --depth=1 --filter=tree:0 -b $BRANCH $FORK_URL
cd $REPO_NAME
git sparse-checkout set --no-cone "extensions/$EXTENSION_NAME"
git checkout
cd "extensions/$EXTENSION_NAME"
npm install && npm run dev

We're currently experiencing a high volume of incoming requests. As a result, the initial review may take up to 15 business days.

[greptile-apps]

Confidence Score: 4/5

[Medium impact] The change affects iMessage code discovery in the 2FA Code Finder.

The earlier missing-code issue remains partly unfixed and should be addressed before merging.

Summary

Restores iMessage code detection when macOS stores message text only in attributedBody.

• The code finder reads message bodies stored in attributedBody.
• The extension notes the fix and credits its contributor.

_{Reviews (2) · Last reviewed commit: "fix(imessage-2fa): limit results after f..." · Reviewed by Greptile}