Send hCaptcha API response data to event platform
Closed, ResolvedPublic
Actions

Description

Summary

Record part of hCaptcha API response data to the Wikimedia Event Platform during account creation

Background

hCaptcha response data risk signal should be stored in the user's global session
The normed risk signal can be retrieved in extensions/Campaigns/includes/CampaignsSecondaryAuthenticationProvider.php and included in in ServerSideAccountCreation event data.
the ServerSideAccountCreation event must opt out of sending IP and user-agent data.

User story

As staff, I want to send hCaptcha response data to the Event Platform during account creation to support analytics, while excluding IP and user-agent from this event stream.

Acceptance criteria

hCaptcha risk signal stored in global session
risk signal included in ServerSideAccountCreation event

Details

Other Assignee: kostajh

Related Changes in Gerrit:

Subject	Repo	Branch	Lines +/-
hCaptcha: Set a default value for remoteip	mediawiki/extensions/ConfirmEdit	master	+13 -4
hCaptcha: Set a default value for remoteip	mediawiki/extensions/ConfirmEdit	wmf/1.45.0-wmf.16	+13 -4
hCaptcha: Enable processing of the risk score	operations/mediawiki-config	master	+3 -0
HCaptcha: Save hCaptcha score in global session	mediawiki/extensions/ConfirmEdit	master	+5 -0
CampaignsSecondaryAuthenticationProvider: Capture hCaptcha-score and include in ServerSideAccountCreation event	mediawiki/extensions/Campaigns	master	+12 -0

Customize query in gerrit

Related Changes in GitLab:

	Title	Reference	Author	Source Branch	Dest Branch
	Add hCaptchaScore and create version 1.4.0 of analytics/legacy/serversideaccountcreation	repos/data-engineering/schemas-event-secondary!49	reedy	1.4.0	master

Customize query in GitLab

Related Objects
Search...

		Status	Subtype	Assigned	Task
		Resolved		Reedy	T379179 Send hCaptcha API response data to event platform
		Resolved		• nettrom_WMF	T397794 Data QA for instrumentation of hCaptcha API response

Event Timeline

Reedy created this task.Nov 6 2024, 5:23 PM

Restricted Application added a project: Data-Engineering. · View Herald TranscriptNov 6 2024, 5:23 PM

Restricted Application added a subscriber: Aklapper. · View Herald Transcript

Ottomata added projects: Test Kitchen, Metrics Platform.Nov 6 2024, 5:27 PM

Ottomata added a subscriber: phuedx.

Ottomata subscribed.

Ahoelzl moved this task from Incoming (new tickets) to Tag with Radar on the Data-Engineering board.Nov 6 2024, 9:22 PM

Reedy moved this task from Backlog to Feature Requests/Improvements on the ConfirmEdit (CAPTCHA extension) board.Nov 6 2024, 11:34 PM

@VirginiaPoundstone: This seems like a good candidate for a tightly-scoped instrument using either the JS or PHP Metrics Platform Client. What do you think?

@phuedx could be. I need a little more context. Let's talk about it during grooming.

• acooper added a project: WE4.2 Bot detection.Dec 13 2024, 2:27 PM

We need to do something like this schema: https://schema.wikimedia.org/repositories/secondary/jsonschema/analytics/mediawiki/ip_reputation/score/current.yaml

@acooper, what would be the producer of the captcha events? Is it code we write and own? If so, this should be easy. If not, we have ways, but it kind of depends on the structure of the events sent.

In T379179#10406142, @Ottomata wrote:

@acooper, what would be the producer of the captcha events? Is it code we write and own? If so, this should be easy. If not, we have ways, but it kind of depends on the structure of the events sent.

Yes, it is code we manage in ConfirmEdit extension. The idea is that in the PageSaveCompleteHook, we'd log an event that also includes the hCaptcha score (if any) for a given request. We'd want to save performer ID, revision ID and bucketed score (not the raw score). For account creations, we'd omit revision ID.

• VirginiaPoundstone moved this task from NEEDS ESTIMATION to Move to Data Engineering on the Test Kitchen board.Jan 6 2025, 8:26 PM

• VirginiaPoundstone removed a project: Test Kitchen.Jan 6 2025, 8:50 PM

• VirginiaPoundstone added a project: Data-Engineering-Radar.Jan 6 2025, 9:10 PM

phuedx moved this task from Backlog to Instruments on the Metrics Platform board.Jan 17 2025, 5:02 PM

• VirginiaPoundstone removed a project: Metrics Platform.Jan 17 2025, 8:58 PM

@Reedy @acooper In extensions/ConfirmEdit/includes/hCaptcha/HCaptcha.php, you should be able to set the normed score in the user's global session (e.g. SessionManager::getGlobalSession()->set( 'hCaptcha', $normedScore );), and then retrieve it in Extension:Campaigns (extensions/Campaigns/includes/CampaignsSecondaryAuthenticationProvider.php) to include it in the event data sent to ServerSideAccountCreation. As noted by @nettrom_WMF, we'll need to update the code logging to ServerSideAccountCreation to opt out of including IP and user-agent data. We want both IP and user agent data for analysis, but we can use MediaWiki-extensions-IPReputation's event data on account creation for that.

Change #1127937 had a related patch set uploaded (by Reedy; author: Reedy):

[mediawiki/extensions/ConfirmEdit@master] HCaptcha: Save hCaptcha score in global session

https://gerrit.wikimedia.org/r/1127937

gerritbot added a project: Patch-For-Review.Mar 14 2025, 4:57 PM

Change #1127940 had a related patch set uploaded (by Reedy; author: Reedy):

[mediawiki/extensions/Campaigns@master] CampaignsSecondaryAuthenticationProvider: Captcha hCaptcha-score and include in ServerSideAccountCreation event

https://gerrit.wikimedia.org/r/1127940

• acooper edited projects, added WE4.2 Bot detection (WE4.2 hCaptcha account creation trial); removed WE4.2 Bot detection.Mar 19 2025, 2:37 PM

reedy opened https://gitlab.wikimedia.org/repos/data-engineering/schemas-event-secondary/-/merge_requests/49

Draft: Add hCaptchaScore and create version 1.4.0 of analytics/legacy/serversideaccountcreation

• acooper renamed this task from Send captcha API response data to event logging to Send hCaptcha API response data to event platform.Mar 21 2025, 2:35 PM

• acooper changed the task status from Open to In Progress.

• acooper triaged this task as Medium priority.

kostajh assigned this task to Reedy.Mar 24 2025, 11:44 AM

kostajh moved this task from Backlog to In progress on the WE4.2 Bot detection (WE4.2 hCaptcha account creation trial) board.

Reedy added a project: MediaWiki-extensions-Campaigns.Mar 24 2025, 3:20 PM

Reedy mentioned this in T389819: Remove deprecated parameters from ServerSideAccountCreation.

kharlan merged https://gitlab.wikimedia.org/repos/data-engineering/schemas-event-secondary/-/merge_requests/49

Add hCaptchaScore and create version 1.4.0 of analytics/legacy/serversideaccountcreation

Change #1127940 merged by jenkins-bot:

[mediawiki/extensions/Campaigns@master] CampaignsSecondaryAuthenticationProvider: Capture hCaptcha-score and include in ServerSideAccountCreation event

https://gerrit.wikimedia.org/r/1127940

ReleaseTaggerBot added a project: MW-1.45-notes (1.45.0-wmf.1; 2025-05-13).May 6 2025, 7:00 AM

kostajh moved this task from In progress to QA on the WE4.2 Bot detection (WE4.2 hCaptcha account creation trial) board.May 8 2025, 2:54 PM

Change #1127937 merged by jenkins-bot:

[mediawiki/extensions/ConfirmEdit@master] HCaptcha: Save hCaptcha score in global session

https://gerrit.wikimedia.org/r/1127937

Maintenance_bot removed a project: Patch-For-Review.May 8 2025, 3:31 PM

kostajh added a project: Trust and Safety Product Sprint (Sprint Carrot Cake (May 26 - June 13)).May 19 2025, 10:38 AM

kostajh moved this task from Priority Backlog to Needs QA on the Trust and Safety Product Sprint (Sprint Carrot Cake (May 26 - June 13)) board.

kostajh edited projects, added Trust and Safety Product Sprint (Sprint Baklava (June 16 - July 4)); removed Trust and Safety Product Sprint (Sprint Carrot Cake (May 26 - June 13)).Jun 16 2025, 6:12 AM

kostajh moved this task from Priority Backlog to Needs QA on the Trust and Safety Product Sprint (Sprint Baklava (June 16 - July 4)) board.

phuedx moved this task from Backlog to Tracking on the MediaWiki-extensions-EventLogging board.Jun 20 2025, 2:23 PM

• sguebo_WMF updated the task description. (Show Details)Jun 24 2025, 11:23 PM

• sguebo_WMF removed a subscriber: • acooper.

• sguebo_WMF mentioned this in T397794: Data QA for instrumentation of hCaptcha API response.Jun 25 2025, 12:26 AM

• sguebo_WMF added a subtask: T397794: Data QA for instrumentation of hCaptcha API response.

• sguebo_WMF changed the task status from In Progress to Stalled.Jul 5 2025, 1:49 AM

• sguebo_WMF moved this task from QA to Backlog on the WE4.2 Bot detection (WE4.2 hCaptcha account creation trial) board.

kostajh edited projects, added Trust and Safety Product Sprint (Sprint Cannoli (July 7 - July 25)); removed Trust and Safety Product Sprint (Sprint Baklava (June 16 - July 4)).Jul 7 2025, 9:42 AM

kostajh moved this task from Priority Backlog to Needs QA on the Trust and Safety Product Sprint (Sprint Cannoli (July 7 - July 25)) board.

This is pending some discussions with hCaptcha before we can re-enable.

kostajh updated the task description. (Show Details)Jul 16 2025, 1:14 PM

kostajh removed a subscriber: • VirginiaPoundstone.

In T379179#10636881, @kostajh wrote:

@Reedy @acooper In extensions/ConfirmEdit/includes/hCaptcha/HCaptcha.php, you should be able to set the normed score in the user's global session (e.g. SessionManager::getGlobalSession()->set( 'hCaptcha', $normedScore );), and then retrieve it in Extension:Campaigns (extensions/Campaigns/includes/CampaignsSecondaryAuthenticationProvider.php) to include it in the event data sent to ServerSideAccountCreation. As noted by @nettrom_WMF, we'll need to update the code logging to ServerSideAccountCreation to opt out of including IP and user-agent data. We want both IP and user agent data for analysis, but we can use MediaWiki-extensions-IPReputation's event data on account creation for that.

I don't remember why we said we needed to opt out of including IP and user-agent data. I think we should leave this as is.

In T379179#11112619, @kostajh wrote:

I don't remember why we said we needed to opt out of including IP and user-agent data. I think we should leave this as is.

I think that's a remnant from our early conversations about what the associated risk tier would be and whether we were aiming for low risk to reduce delays from having to go through full L3SC review. Please do correct me if I'm wrong, but I seem to remember that we've been through L3SC review now and gotten signoff on the associated risk, and therefore we don't need to make any changes.

Change #1182824 had a related patch set uploaded (by Kosta Harlan; author: Kosta Harlan):

[operations/mediawiki-config@master] hCaptcha: Enable processing of the risk score

https://gerrit.wikimedia.org/r/1182824

gerritbot added a project: Patch-For-Review.Aug 28 2025, 12:31 PM

kostajh changed the task status from Stalled to In Progress.Aug 28 2025, 12:31 PM

kostajh updated Other Assignee, added: kostajh.

kostajh removed a subscriber: • acooper.

Change #1182824 merged by jenkins-bot:

[operations/mediawiki-config@master] hCaptcha: Enable processing of the risk score

https://gerrit.wikimedia.org/r/1182824

Mentioned in SAL (#wikimedia-operations) [2025-08-28T12:35:09Z] <kharlan@deploy1003> Started scap sync-world: Backport for [[gerrit:1182824|hCaptcha: Enable processing of the risk score (T379179)]]

Mentioned in SAL (#wikimedia-operations) [2025-08-28T12:41:48Z] <kharlan@deploy1003> kharlan: Backport for [[gerrit:1182824|hCaptcha: Enable processing of the risk score (T379179)]] synced to the testservers (see https://wikitech.wikimedia.org/wiki/Mwdebug). Changes can now be verified there.

Mentioned in SAL (#wikimedia-operations) [2025-08-28T12:49:22Z] <kharlan@deploy1003> Finished scap sync-world: Backport for [[gerrit:1182824|hCaptcha: Enable processing of the risk score (T379179)]] (duration: 14m 13s)

Maintenance_bot removed a project: Patch-For-Review.Aug 28 2025, 1:31 PM

Change #1182844 had a related patch set uploaded (by Kosta Harlan; author: Kosta Harlan):

[mediawiki/extensions/ConfirmEdit@master] hCaptcha: Set a default value for remoteip

https://gerrit.wikimedia.org/r/1182844

gerritbot added a project: Patch-For-Review.Aug 28 2025, 2:14 PM

kostajh added a project: Trust and Safety Product Sprint (Sprint Princess Tarta (August 18 - September 5)).Aug 28 2025, 2:48 PM

kostajh moved this task from Priority Backlog to Needs Review on the Trust and Safety Product Sprint (Sprint Princess Tarta (August 18 - September 5)) board.

Change #1182866 had a related patch set uploaded (by Kosta Harlan; author: Kosta Harlan):

[mediawiki/extensions/ConfirmEdit@wmf/1.45.0-wmf.16] hCaptcha: Set a default value for remoteip

https://gerrit.wikimedia.org/r/1182866

Change #1182866 merged by jenkins-bot:

[mediawiki/extensions/ConfirmEdit@wmf/1.45.0-wmf.16] hCaptcha: Set a default value for remoteip

https://gerrit.wikimedia.org/r/1182866

Mentioned in SAL (#wikimedia-operations) [2025-08-28T15:30:53Z] <kharlan@deploy1003> Started scap sync-world: Backport for [[gerrit:1182866|hCaptcha: Set a default value for remoteip (T379179)]]

Mentioned in SAL (#wikimedia-operations) [2025-08-28T15:34:56Z] <kharlan@deploy1003> kharlan: Backport for [[gerrit:1182866|hCaptcha: Set a default value for remoteip (T379179)]] synced to the testservers (see https://wikitech.wikimedia.org/wiki/Mwdebug). Changes can now be verified there.

Change #1182844 merged by jenkins-bot:

[mediawiki/extensions/ConfirmEdit@master] hCaptcha: Set a default value for remoteip

https://gerrit.wikimedia.org/r/1182844

Mentioned in SAL (#wikimedia-operations) [2025-08-28T15:41:17Z] <kharlan@deploy1003> Finished scap sync-world: Backport for [[gerrit:1182866|hCaptcha: Set a default value for remoteip (T379179)]] (duration: 10m 24s)