iOS parameterized input remains in native XCTest diagnostic logs
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 4/5
- Tempo stimato
- 3-5 giorni
- Idoneità per principianti
- 36/100
- Tipo di issue
- Bug
- Chiarezza
- Abbastanza chiara
- Stato di attività
- Attiva
- Stack tecnologico
- ios, typescript
- Ambito
- mobile-dev, security
Direzione di ricerca
Start by tracing the Node API entry point client.interactions.fill and how input reaches the stock XCTest runner; inspect the .xcresult diagnostic log behavior described in the reproduction. Done means establishing whether a supported stock path prevents sensitive values from appearing there, or documenting the limitation if it cannot.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Observed behavior
On unmodified [email protected], a harmless value supplied through the Node API with fill({ text, recordAs }) was correctly absent from the fill response, subsequent predicate/error responses, and the inspected Agent Device session artifacts, but remained verbatim in the native XCTest .xcresult diagnostic log.
We are trying to use the official release for iOS login automation without carrying a downstream patch. Is there a supported way to prevent sensitive visible input from being retained by this native diagnostic channel? If this is outside the parameterization guarantee, documenting that boundary would also help.
Environment
- Official npm release 0.21.22; source tag commit
202585240f0c1289642885b5df645cbc25114272 - macOS host, Xcode 27.0 (27A266a), iOS Simulator
- Stock signed XCTest runner; no modified upstream code
- Script recording armed for
recordAs; no screen recording, trace, or app-log capture requested - Only inert test strings were used, never a real credential
Reproduction outline
-
Boot an owned iOS simulator and open Settings with the stock CLI in a fresh state directory/session:
agent-device open com.apple.Preferences --platform ios --udid "$SIM_UDID" \ --session stock-canary --save-script --foreground --timeout 120000 --json -
Use the official Node client for that same session. Take a fresh snapshot and select the unique Settings
SearchFieldwith labelSearch. Then execute:const options = { platform: 'ios', udid: process.env.SIM_UDID }; const client = createAgentDeviceClient({ session: 'stock-canary', cwd: process.cwd() }); // searchRef comes from the fresh Settings snapshot. await client.interactions.fill({ ...options, ref: searchRef, text: 'INERT_STOCK_PRIVATE_CANARY_20261006', recordAs: 'PRIVATE_CANARY', }); await client.interactions.is({ ...options, predicate: 'exists', selector: 'label="Settings"' }); try { await client.interactions.press({ ...options, ref: '@e999999' }); } catch {} -
Inspect the exact runner's
.xcresultStaging/1_Test/Diagnostics/.../Session-AgentDeviceRunnerUITests-*.logfor the inert marker. Our observed excerpt:SearchField, label: 'Search', Keyboard Focused ... faulted in { "XC_kAXXCAttributeValue" = "INERT_STOCK_PRIVATE_CANARY_20261006"; }: (no error) for value using XC_kAXXCAttributeValue
A second run using another inert marker and fill({ ..., verify: false }) also retained its marker in the native diagnostic log. The stock XCTest plan already sets system/user attachment lifetime to keepNever; that did not prevent this diagnostic text.
This establishes persistence during the sequence, not that the deliberately failed press alone caused it. We have not yet established the behavior on a physical iPhone or a masked password field. The visible Settings input was a control for a visible OTP field.
Expected / requested guidance
A documented stock path for sensitive visible input that avoids retaining its literal value in native diagnostics, or an explicit documented limitation so callers can stop before supplying credentials. We are not asking to disable ordinary observations globally.
Related: #1348 and #1398 provide useful recording/response redaction, and those protections worked in the checks above. #3260 separately concerns CLI input transport; it would not by itself address native diagnostics.
- Lingua principale
- TypeScript
- Stelle
- 4.9k
- Fork
- 328
- Merge medio
- 12h 13m
- PR unite (30g)
- 541
Preparare l'ambiente
- Nessun Dockerfile né file Docker Compose
- Nessun modello di pull request
- Leggi la guida per i contributori
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di callstack/agent-device
-
settings clear-app-state fails with ENOENT … scandir '(null)' for iOS system apps such as SettingsForse già presa @thymikee l’ha presa oggi. Aperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 80/100
callstack/agent-device#3305 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 68/100
callstack/agent-device#1869 ·
I maintainer di solito rispondono entro 1 giorno
-
install and reinstall from another session in the same daemon replace the app on a claimed deviceAperta
Difficoltà 4/5 3-5 giorni Idoneità per principianti 55/100
callstack/agent-device#3345 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 3-5 giorni Idoneità per principianti 48/100
callstack/agent-device#3344 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 5/5 Più di una settimana Idoneità per principianti 35/100
callstack/agent-device#3342 · 7 commenti ·
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di callstack/agent-device
Issue simili
-
bug priority:low ready-for-dev
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 70/100
Automattic/data-liberation-agent#685 ·
I maintainer di solito rispondono entro 1 giorno
-
Business
Difficoltà 2/5 1-3 ore Idoneità per principianti 66/100
I maintainer di solito rispondono entro 1 giorno
-
bug
Difficoltà 2/5 1-3 ore Idoneità per principianti 83/100
txn2/mcp-data-platform#2063 ·
I maintainer di solito rispondono entro 1 giorno
-
bug
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 77/100
Crosstalk-Solutions/project-nomad#1427 ·
I maintainer di solito rispondono entro 2 giorni