fix: deep link to world redirects to Genesis after account change

[Maurizio-dcl]

Pull Request Description

Fix #8802

What does this PR change?

This PR moves the world access check from bootstrap to post-auth in the loading flow, this fixes a bug where changing accounts would lead to an inconsistent state regarding world permission since they were checked only against the initial identity.

Moved world access check from bootstrap to post-auth in the loading flow, so it runs against the identity the user actually picked (rather than the cached one).
Removed the now-dead IsUserAuthorisedToAccessWorldAsync from IRealmController.

Test Instructions

Prerequisites

• Launch this PR's build with --realm italy2026.dcl.eth --position 20,20 --debug args.
• Two accounts available: one with access to italy2026.dcl.eth, one without (can be tested with any world to which at least one of your accounts have access).

Test Steps

1. Primary repro (cached account has NO access, switch to one that does):

   • Make sure the cached account does NOT have access to the target restricted world
   • Open the client via deep link to the restricted world
   • On the auth screen, click "Change Account" and log in with the account that DOES have access
   • Click "Enter Decentraland"
   • Expected: lands in the world (no Genesis fallback)

2. Reverse case (cached account HAS access, switch to one that doesn't):

   • Cached account has access to the restricted world
   • Open the client via deep link to that world
   • On the auth screen, change to an account without access
   • Click "Enter Decentraland"
   • Expected: loading screen shows, then the user is sent to Genesis with a chat message "Could not enter '' due to world permissions. You were sent to Genesis Plaza."

3. Direct entry regression:

   • Cached account has access → deep link to that world → enter without switching accounts
   • Expected: lands in the world at the deep-linked coordinates

4. Public world:

   • Deep link to a public/unrestricted world → enter with any account
   • Expected: lands in the world

5. Normal launch (no deep link):

   • Expected: goes to home/Genesis as usual

Quality Checklist

• Changes have been tested locally
• Documentation has been updated (if required)
• Performance impact has been considered
• For SDK features: Test scene is included

[github-actions]

Windows and Mac build successful in Unity Cloud! You can find a link to the downloadable artifact below.

Name	Link
Commit	582f123
Logs	https://github.com/decentraland/unity-explorer/actions/runs/26891860539
Download Windows	https://github.com/decentraland/unity-explorer/suites/72159230036/artifacts/7388699637
Download Windows S3	https://explorer-artifacts.decentraland.org/@dcl/unity-explorer/branch/fix/8802-deep-link-into-a-world-takes-to-genesis-city-instead/pr-22706-582f123/Decentraland_windows64.zip
Download Mac	https://github.com/decentraland/unity-explorer/suites/72159230036/artifacts/7388650126
Download Mac S3	https://explorer-artifacts.decentraland.org/@dcl/unity-explorer/branch/fix/8802-deep-link-into-a-world-takes-to-genesis-city-instead/pr-22706-582f123/Decentraland_macos.zip
Built on	2026-06-03T15:15:55Z

[m3taphysics]

PR #8848, run #26229701217

Builds: Windows change, Windows baseline, macOS change, macOS baseline

Framework 13 i7

Metric	Change	Baseline	Delta	Improvement
Samples	2700	2700
CPU average	33.3 ms	33.3 ms	0.0 ms	-0.0%	⚪
CPU 1% worst	33.6 ms	33.5 ms	0.1 ms	-0.3%	⚪
CPU 0.1% worst	35.6 ms	34.5 ms	1.1 ms	-3.3%	🔴
GPU average	7.9 ms	8.1 ms	-0.1 ms	1.7%	⚪
GPU 1% worst	19.2 ms	19.6 ms	-0.4 ms	2.1%	⚪
GPU 0.1% worst	25.2 ms	26.3 ms	-1.0 ms	4.0%	🟢

[m3taphysics]

PR #8848, run #26742207444

Builds: Windows change, Windows baseline, macOS change, macOS baseline

Framework 13 i7

Metric	Change	Baseline	Delta	Improvement
Samples	2700	2699
CPU average	33.3 ms	33.3 ms	-0.0 ms	0.0%	⚪
CPU 1% worst	33.5 ms	34.9 ms	-1.4 ms	4.0%	🟢
CPU 0.1% worst	34.9 ms	49.2 ms	-14.3 ms	29.0%	🟢
GPU average	8.0 ms	7.2 ms	0.8 ms	-10.8%	🔴
GPU 1% worst	17.2 ms	19.2 ms	-2.0 ms	10.2%	🟢
GPU 0.1% worst	19.8 ms	24.1 ms	-4.4 ms	18.1%	🟢

[sandrade-dcl]

LGTM! I only left a comment about a doubt that I would like you to take a look, please. Thanks!

[m3taphysics]

PR #8848, run #26808743981

Builds: Windows change, Windows baseline, macOS change, macOS baseline

Framework 13 i7

Values are the median across 3 run(s); the (min–max) range shows how much the runs varied. A metric is only flagged when the change is larger than both the 3% threshold and that run-to-run variation.

Metric	Change	Baseline	Delta	Improvement
Samples	2700 (×3)	2700 (×3)
CPU average	33.3 ms	33.3 ms	0.0 ms	-0.0%	⚪
CPU 1% worst	34.1 ms (33.5–34.1)	33.8 ms (33.4–33.8)	0.3 ms	-0.9%	⚪
CPU 0.1% worst	39.2 ms (33.6–40.2)	36.5 ms (33.6–39.1)	2.8 ms	-7.6%	⚪
GPU average	7.7 ms (7.5–7.8)	9.5 ms (8.4–10.0)	-1.7 ms	18.1%	🟢
GPU 1% worst	18.8 ms (17.9–19.8)	18.8 ms (18.5–19.4)	0.0 ms	-0.0%	⚪
GPU 0.1% worst	23.1 ms (22.7–23.9)	25.9 ms (25.9–26.0)	-2.8 ms	10.9%	🟢
Exceptions logged	22	18	+4		🔴

[m3taphysics]

PR #8848, run #26812819865

Builds: Windows change, Windows baseline, macOS change, macOS baseline

Framework 13 i7

Values are the median across 3 run(s); the (min–max) range shows how much the runs varied. A metric is only flagged when the change is larger than both the 3% threshold and that run-to-run variation.

Metric	Change	Baseline	Delta	Improvement
Samples	2700 (×3)	2700 (×3)
CPU average	33.3 ms	33.3 ms	0.0 ms	-0.0%	⚪
CPU 1% worst	33.8 ms (33.8–33.9)	33.9 ms (33.5–34.1)	-0.1 ms	0.3%	⚪
CPU 0.1% worst	36.6 ms (36.5–37.5)	36.3 ms (35.1–39.6)	0.3 ms	-0.7%	⚪
GPU average	7.7 ms (7.2–8.3)	8.3 ms (7.3–8.3)	-0.6 ms	7.3%	⚪
GPU 1% worst	19.5 ms (16.8–21.0)	19.9 ms (17.5–20.6)	-0.5 ms	2.3%	⚪
GPU 0.1% worst	25.0 ms (20.8–25.6)	25.0 ms (24.0–25.5)	-0.0 ms	0.2%	⚪
Exceptions logged	18	14	+4		🔴

[claude]

Re-reviewing after latest commits

• Read changed files
• Check if previous blocking issues were resolved
• Review latest changes
• Post final review

View job run

[claude]

Claude finished @Maurizio-dcl's task in 1m 45s —— View job

---

Re-reviewing after latest commits

• Read changed files
• Check if previous blocking issues were resolved
• Review latest changes
• Post final review

---

Root-Cause Assessment

PASS — The root cause (world access checked at bootstrap against the cached/pre-auth identity) is correctly fixed by deferring the check to VerifyWorldAccessAndFallbackIfNeededAsync inside ShowWhileExecuteTaskAsync, after the user has confirmed their identity. This is not a workaround.

---

Both Previous Blocking Issues: RESOLVED ✅

Issue 1: IRealmController.Fake.IsUserAuthorisedToAccessWorldAsync
Removed in commit 582f123b. The Fake nested class no longer contains this stub. ✅

Issue 2: RealmController.TryExtractWorldName
Removed in commit 2cd36d7f. The private static method and all its dead dependencies (IWeb3IdentityCache, IWorldPermissionsService fields) are gone from RealmController. ✅

OperationCanceledException handling is correct — catch (OperationCanceledException) { return; } silently absorbs cancellation (commit 5539107e/19f845a2). ✅

---

No New Blocking Issues

The final shape of VerifyWorldAccessAndFallbackIfNeededAsync (RealUserInAppInitializationFlow.cs:239) is clean:

• Guard clauses skip the check for local scene development, non-world realms, and null domain.
• OperationCanceledException is filtered, not rethrown.
• Network failure during the check triggers Genesis fallback (defensive policy, intentional).
• PasswordRequired also falls back to Genesis — correct for the startup/deep-link path (the interactive password prompt is handled elsewhere via PrivateWorldsPlugin).

---

Non-Blocking Warnings

+4 exceptions logged — The performance bot reports +4 exceptions vs. baseline in the two most recent runs (normal Genesis playthrough, no deep link). These may be unrelated to the auth-flow path changed here. Worth confirming before merge.

Empty world name in fallback message — When TryExtractWorldName returns false, worldName is string.Empty and the chat message reads: "Could not enter '' due to world permissions. You were sent to Genesis Plaza." Consider falling back to the raw domain string for a better user message. Non-blocking.

---

REVIEW_RESULT: PASS ✅
COMPLEXITY: COMPLEX
COMPLEXITY_REASON: Modifies auth flow, world permissions service integration, async UniTask patterns, plugin/container wiring, and the initialization flow sequence.
QA_REQUIRED: YES

[github-actions]

🔍 Claude reviewed this PR and found no blocking issues, but assessed it as complex — human DEV review is still required before merging.

[DafGreco]

✔️ PR reviewed and approved by QA on both platforms following instructions playing both happy and un-happy path

Regressions for this ticket had been performed in order to verify that the normal flow is working as expected:

• [✔️ ] Log In/Log Out
• [✔️ ] Backpack and wearables in world
• [ ✔️] Emotes in world and in backpack
• [ ✔️] Teleport with map/coordinates/Jump In
• [ ✔️] Chat and multiplayer
• [ ✔️] Settings

⚠️ Important comment ⚠️
We found that after loggin in with an account with permission and then logging out and loggin in with an account without permissions does not trigger any message on the chat regarding to not having permissions to enter the world
This has been a topic that is found on prd and this PR so in any case needed we can create a ticket later

Evidence:

20260603-1249-28.7951655.mp4

[m3taphysics]

PR #8848, run #26894354241

Builds: Windows change, Windows baseline, macOS change, macOS baseline

How to read this table

• Each build is measured 3 times. The values are the median, and (min–max) is the lowest and highest of those runs — a wide range means the metric is noisy and small differences are not trustworthy.
• Δ is Change minus Baseline (a negative Δ means Change is faster).
• 🟢 faster / 🔴 slower — a real difference: larger than both 3% and the run-to-run range.
• ⚪ within noise — the difference is smaller than how much the build varies between its own runs, so it cannot be told apart from random variation. Treat it as no change.
• Exceptions logged — exceptions found in the run logs; more than the baseline is flagged 🔴 even when frame times look fine.

Framework 13 i7

Metric	Baseline	Change	Δ	Result
Samples	2701 (×3)	2700 (×3)
CPU average	33.3 ms	33.3 ms	-0.0 ms	⚪ within noise
CPU 1% worst	34.4 ms (33.9–34.7)	33.7 ms (33.5–33.8)	-0.7 ms	⚪ within noise
CPU 0.1% worst	43.3 ms (37.5–45.2)	35.7 ms (34.7–36.0)	-7.5 ms	⚪ within noise
GPU average	7.3 ms (7.3–7.4)	8.2 ms (8.2–8.6)	0.9 ms	🔴 12% slower
GPU 1% worst	19.6 ms (18.1–20.2)	18.0 ms (17.1–18.1)	-1.6 ms	⚪ within noise
GPU 0.1% worst	22.9 ms (22.8–28.9)	23.9 ms (21.4–25.2)	0.9 ms	⚪ within noise
Exceptions logged	46	18	-28	🟢 fewer errors