Sprint - Feb 9 to Feb 22

[pl]

Sprint Planning

List of incidents + any action items that haven't been scheduled yet

Team sprint planning

For your team sprint planning copy this template into a comment below for each team.

# Team ___

**Support hero:** ___
**Off during the sprint:** ___

## Quarter goals

1. Quarter goals, status (red/yellow/green) and action points if yellow/red

## Retro

### Tasks

<!-- Grab the high and low priority items from last time and add whether that item was completed or not -->

## Plan

### High priority

-

### Low priority / side quests

-

### Are any other teams impacted by this plan? If so, tag them here

-

[pl]

Team Ingestion

• Support: @nickbest-ph

Retro

Details

High priority

• Persons DB / API
  • 🟢 Drop unused indexes @nickbest-ph
  • 🟢 Finalize the RFC @nickbest-ph
  • 🟢 PoC the Persons Read API @nickbest-ph
  • 🔴 PoC for the Persons Merge API logic with mocked dependencies @pl
• Deduplicator
  • 🟢 Get checkpoints to work @eli-r-ph
  • 🟡 Implement the catch-up based on the output topic @eli-r-ph
  • 🔴 Dry run on production @eli-r-ph @jose-sequeira
• LLM Analytics Ingestion
  • 🟡 Implement request signature validation @pl
  • 🟡 Revise the placeholders in the output event format @pl
• Event Restrictions
  • 🟢 Implement AND logic for the event restriction filters @pl
  • 🟢 Deploy the dynamic restrictions in capture @pl
• Ingestion Platform
  • 🟢 Remove legacy ingestion consumer pipeline code @pl
  • 🟢 Document the pipelines framework @pl
  • 🔴 Prepare the ingestion pipeline for WS testing @lmenezes
  • 🟢 Stretch: Start moving Blobby to pipelines @pl
  • 🟢 Clean up Kafka topics @pl
• Rate Limiter
  • 🟡 Test the PoC on capture-mirror @eli-r-ph
  • 🔴 Stretch: do a dry run on capture @eli-r-ph
  • 🟡 Finish the temporal job for distinct ID misuse reports @jose-sequeira
• 🟢 Finish separating the Node.js builds @jose-sequeira
• 🟢 Run test harness reports in CI @jose-sequeira
• 🟢 Leo's onboarding @pl @jose-sequeira @lmenezes

Incident follow-up

• 🟡 Write a public post-mortem for the person lost properties incident @pl
• 🟡 Implement stateful overflow @jose-sequeira

Support follow-up

• 🟢 Removing property definitions @jose-sequeira
• 🔴 Drop $ip property before transformations @pl
• 🟡 Persons without distinct IDs in CH @pl

Side quests

None

OKRs

• 🟢 Fence off non-ingestion services from persons DB
• 🟢 Introduce rate / usage limits
• 🟢 Move Error Tracking ingestion to ingestion platform
• 🟢 Implement core features for the ingestion platform

High priority

• Persons DB / API
  • Figure out how to test feature flags integration on prod @nickbest-ph
  • Implement the API client for feature flags @nickbest-ph
  • PoC for the Persons Merge API logic with mocked dependencies @pl
• Deduplicator
  • Prepare the deduplicator to be deployed behind ingestion @jose-sequeira
  • Investigate running the deduplicator ahead of ingestion-events-historical @jose-sequeira @eli-r-ph
  • Implement a fail-open flag @eli-r-ph @jose-sequeira
  • Implement the catch-up based on the output topic @eli-r-ph
• LLM Analytics Ingestion
  • Implement request signature validation @pl
  • Revise the placeholders in the output event format @pl
• Ingestion Platform
  • Prepare the ingestion pipeline for WS testing @lmenezes
  • Stretch: clean up filterOkMap steps @pl
• Session Replay Ingestion
  • Switch blobby to the new topics @pl
• Rate Limiter
  • Test the PoC on capture-mirror @eli-r-ph
  • Stretch: do a dry run on capture @eli-r-ph
  • Finish the temporal job for distinct ID misuse reports @jose-sequeira
• @z0br0wn's onboarding @pl @jose-sequeira

Incident follow-up

• Test stateful overflow on prod @jose-sequeira
• Implement acceptance tests @lmenezes

Support follow-up

• Drop $ip property before transformations @pl
• Persons without distinct IDs in CH @pl
• Investigate event duplication on rebalancing @jose-sequeira
• Investigate duplication in Mixpanel managed migrations @nickbest-ph

Side quests

TBD

[ioannisj]

Team Mobile

Support hero: @ioannisj
OOO:

Quarter goals

• 🟡 Goal 1: Replay cost control
• 🟡 Goal 2: Mobile error tracking v2
• ⚪ Goal 3: Wider SDK coverage
• 🟡 Goal 4: Make it easier to run the team

Retro

• shifted RN manual recording to high priority after requests @ioannisj
• finally client libraries team yay @marandaneto
• Dustin joins the team @marandaneto

Details

High priority

• ⚪ React Native plugin for native exceptions - error tracking (Android and iOS) @marandaneto
  • still blocked by iOS bits
• 🟡 RN session replay plugin release fully automated and part of the monorepo @marandaneto
• 🟡 Apple error tracking support @ioannisj
• 🔴 iOS - Session Replay feature degrades performance @ioannisj
• 🟡 Some pending/draft PRs from previous sprint that need to be merged @ioannisj
• 🟡 RN - Manually start and stop session recordings @ioannisj
• 🟢 mobile team -> client libraries team changes @marandaneto

Low priority / side quests

• 🟢 Android - Attach person and group properties to /flags request @marandaneto
• 🟢 RN - Capture $set event when identify is called with the same distinct id and different user props @marandaneto
• 🟢 RN - setPersonProperties support ($set event) @marandaneto
• ⚪ PostHogMaskWidget not masking properly on wide screens @ioannisj
• 🟡 fix queue pending feature flags reload instead of dropping requests @marandaneto @ioannisj
  • 🟢 Android @marandaneto
  • 🟢 RN @marandaneto
  • iOS

Plan

High priority

• iOS - Session Replay feature degrades performance @ioannisj
• Apple error tracking support - release + docs @ioannisj
• React Native plugin for native exceptions - error tracking (Android and iOS) @marandaneto
  • still depends on iOS unblocking this
• iOS - queue pending feature flags reload instead of dropping requests @ioannisj
• Enable/Disable autocapture via remote config - SDK, part of this @marandaneto
  • Android @marandaneto
  • RN @marandaneto
  • iOS
  • Flutter
• Error tracking user interviews (2?) @marandaneto
• Web SDK bundle size reductions @dustinbyrne
• Flutter / React / React Native support for getFeatureFlagResult API + docs @dustinbyrne

Low priority / side quests

• ⚪ PostHogMaskWidget not masking properly on wide screens @ioannisj
  • I'll see if I can spot the last edge case reported by customer, otherwise I'll release what I have which should cover most of the cases anyway. This only happens when using custom designSize in flutter screen utils.
• Refactor iOS PostHogSDK class @ioannisj
• iOS: Convert all unit tests to use swift testing framework @ioannisj
• Use new PostHogMulticastCallback across iOS codebase @ioannisj
• Session replay FE changes for remote config changes on iOS @ioannisj

🟢 =finished 🟡=in progress 🔴=won't finish ⚪=not started

[rossgray]

Team Batch Exports

Support hero: @rossgray
Off during the sprint: @tomasfarias

Quarter goals

• Objective 1: Export anything anywhere
• Objective 2: Keep batch exports running into the future

Retro

Details

High priority

• ✅ Daily batch exports improvements @rossgray
  • Customize time
  • Customize timezone
  • Consistent timezone when backfilling
• ✅ Enable weekly batch exports @rossgray
• 🟡 Discuss deprecation of HTTP batch exports (by enabling CDP destinations)
  • ⚪ Disable them for everyone not currently using them
  • 🟡 Figure out what we need to do to enable CDP
  • Waiting on Workflows team to finish infra setup

Low priority

• ⚪ Warning when triggering a backfill (count of estimated rows)

Other things

• Databricks now in open beta
• Setting up product intent for batch exports
• Updated our temporal worker scaling config and memory limits
• Speeding up our Python tests
• Security incident

✅=finished 🟡=in progress 🔴=won't finish ⚪=not started

High priority

• Discuss deprecation of HTTP batch exports (by enabling CDP destinations) @rossgray
  • Disable them for everyone not currently using them
  • Figure out what we need to do to enable CDP
• 🟡 Warning when triggering a backfill (count of estimated rows) @rossgray

Low priority

• ✅ Use passwordless authentication for internal stage queries @rossgray
• Update tooling for manual backfilling (eg due to incidents, customer support) @rossgray
  • Need to support custom start times, timezones, weekly schedule
• 🟡 Introduce some frontend tests @rossgray
• ✅ Look into OOMs @rossgray
  • changed memory limits and autoscaling config
  • didn't manage to find any way to prevent memory leaks

[MichaelKutsch-ph]

Team Infra

Support hero: @bciaraldi

• 🟡 The Application™ @langesven @bciaraldi
  • temporal-workers are in the making
  • it makes sense to move things to namespaces now
• ⚪ cost reduction - if IO optimized is still required for the main DBs @danielxnj @bciaraldi
• ⚪Cilium deployment to dev @danielxnj - swapped for rate limiting
• ✅ rate limiting @danielxnj - added to all envs now - doc, dashboards, alerts in place 🥳
• ✅ Cloudflare reverse proxy opt-in in the UI for users to accept Cloudflare as data processor @danielxnj
  • released as beta, documentation, alerts provided
• ⚪ Granian (or alternative) webserver for Django stack @langesven
• 🟡 Support team data stack with ducklings @MichaelKutsch-ph @bciaraldi
  • this was not so great - I (michi) was not close enough to them plus timezone-gap. this kept iteration speed low and misunderstandings: next time, stay way closer, align on timezones, directly build something that can be reused. Sync upfront for a better planning
  • current state: I think they can mostly run on their own now, but we need to stay close
  • TO DO: map out standard-building blocks and EKS / network architecture that we want to work towards
• 🟡 make billing development flow great @MichaelKutsch-ph
  • revenue-dev created, currently setting up some tooling
• 🟡 loads of terraform-cleanup @MichaelKutsch-ph - shall we move towards heavy scripting + claude-skills as a sort of service catalog for infra?

Low priority / side quests

• vmalert loki not alerting properly - deadmanssnitch @support hero
• local eval: high CPU sage lately - check to optimize nodepools @bciaraldi
• ⚪ AWS Nuke or ephemeral AWS accounts @MichaelKutsch-ph

Retro

✅=finished 🟡=in progress 🔴=won't finish ⚪=not started

OKR Infra

• 🛤️ Golden path for Terraform/Terragrunt - Drivers: @MichaelKutsch-ph @danielxnj
• 🚀 The Application™ - effortless app deployments - Drivers: @langesven @bciaraldi
• 🌐 Replace AWS CNI with Cilium - Drivers: @Piccirello @danielxnj @bciaraldi
• 🎯 Dedicated load balancer for Capture
• 📦 Dependency upgrades (@bciaraldi coordinating, support hero executing)
• [full list of goals]https://github.com/PostHog/posthog-cloud-infra/issues/6088)

High Prio

• The Application™ @langesven @bciaraldi
• Cloudflare reverse proxy tweaks @danielxnj
  • prs approved, going on!
• Cilium deployment to dev @danielxnj
  • spining up new cluster for tests on dev
• RFC cluster architecture - break down those clusters! @MichaelKutsch-ph
• Granian (or alternative) webserver for Django stack @langesven
• make billing development flow great @MichaelKutsch-ph
  • revenue-dev created, currently setting up some tooling
• EKS control plane:
  • clean up unused daemon sets @langesven
  • replace promtail with grafana alloy @langesven
  • adding control plane dashboard
• rate limit tuning necessary @danielxnj
• image clean up @bciaraldi
  security: https://github.com/PostHog/posthog-cloud-infra/issues/6590

[frankh]

Team Logs

Write more bytes somehow

Retro

Details

Main goal: 🚢

• 🟡 Retention options (just for team 2, no pricing yet)
  • 🔴 Retention by service_name @frankh
• 🟢 Logs in GA
  • 🟢 Launch content/comms @abheek9
  • 🟢 Possibly 14 or 15 day retention for launch? @frankh
• 🔴 context view @jonmcwest
• 🟡 saving views/queries @jonmcwest
• better AI analysis @jonmcwest
  • 🟢 MCP and/or PostHog AI integration? @frankh
• 🟡 better handling of json column sub attributes @jonmcwest
• 🔴 loading indicator @jonmcwest
• 🔴 simple JSON endpoint, and curl example command for ingestion
• 🟢 JSON parsing option?? @frankh
• 🟡 breakdown by services @jonmcwest
• 🟢 basic ingestion pipeline stuff: @frankh
  • 🔴 PII/sensitive data scrubber
  • 🟢 JSON parsing
  • 🔴 ??? regex replace?
  • 🔴 dropping logs

High priority

• Retention @frankh
• context view @frankh
• Alerting on logs @jonmcwest
• growth/docs @jonmcwest
• pricing for retention @abheek9
• parallel pipeline for iceberg ingestion? (investigation) @frankh
• logs view model @jonmcwest

Low priority / side quests]

• pipeline stuff
  • 🔴 PII/sensitive data scrubber
  • 🔴 ??? regex replace?
  • 🔴 dropping logs
• hogql
  🟢 =finished 🟡=in progress 🔴=won't finish ⚪=not started

[havenbarnes]

Team Workflows

Off during the sprint: Haven 2/18-2/20

Retro

High priority

• Email sending throttle (per second rate limit for SES) @havenbarnes
• Launch Batch Workflows for Team2 @havenbarnes
  • verify persons timezones is working as expected with Batch Worklows @havenbarnes
• Push notification channel (android) @odin-posthog
  • create docs around person merging & how we use it with push
• cyclotrons rebirth @meikelmosby
  • get shadow writes in place & test new db provider on dev & us with hog queue @meikelmosby
  • come up with simplified version of cyclotron that allows for new features (replay, etc) @meikelmosby

Low priority / side quests

• Allow moving around, repositioning nodes @meikelmosby
• Don't evaluate the next step of a delay action until after the delay has expired @odin-posthog
• make behavioral cohorts work for team 2 (@meikelmosby assisting)
  • Emit cohort_membership_changed events for workflows (shoving data into pg, emitting event for workflows)
• revamp metrics overview for workflow @havenbarnes
• revamp metrics on the node view @havenbarnes

This sprint

High priority

• Launch Batch Workflows for Team2 @havenbarnes
• Launch one-click unsubscribe (configurable by message category) @havenbarnes
• Finish destination backfill work Support Backfills #41634 @havenbarnes
• cyclotrons rebirth @meikelmosby
  • come up with simplified version of cyclotron that allows for new features (replay, etc) @meikelmosby
• make behavioral cohorts work for team 2 (@meikelmosby assisting)
  • Emit cohort_membership_changed events for workflows (shoving data into pg, emitting event for workflows)

Low priority

• revamp metrics overview for workflow @havenbarnes
• revamp metrics on the node view @havenbarnes

[Daesgar]

Team ClickHouse

AKA Team Read Fewer Bytes Please

Retro

Details

Incidents 🔥

#inc-2026-02-02-clickhouse-us-load-spike

High priority

• 🟡 monitor queries missing tags and limits @orian
  • fix experiments using the batch exports users
• 🟡 estimate query "cost" based on query_log_archive @orian
• 🟢 backfill the missing query_log_archive data @orian
• 🟢 switch backups to the new US account @Daesgar
• ⚪ work in a breaking-glass scenario for incidents @Daesgar
• 🟡 make sharded_events_recent consistent across environments @Daesgar
• 🟡 Explore engine to separate S3 and compute with ClickHouse @rorylshanks
• 🟢 Remove property keys from set of events for a request @orian

High priority

• 🟡 make sharded_events_recent consistent across environments @Daesgar
• Enforce privileges to CH users @Daesgar
• Create ops CH instance to start tracking some disk usage statistics @Daesgar
• 🟡 Explore engine to separate S3 and compute with ClickHouse @rorylshanks
• 🟡 monitor queries missing tags and limits @orian
• 🟡 estimate query "cost" based on query_log_archive @orian
• Review and push predicate push-down for HogQL @orian

Med prio

🟢 =finished 🟡=in progress 🔴=won't finish ⚪=not started

Questions / Discussion

[lucasheriques]

Team Surveys Product Experiences

Sprint start: 4 Feb to 18 Feb

• Surveys Support hero: @lucasheriques

🟢=finished 🟡=in progress 🔴=won't finish ⚪=not started

Retro - Team Notes

Retro

What could've been better

• @lucasheriques setting up the experiement for guided surveys took longer than anticipated (key events were missing)

What went well

• guided surveys did great on the experiment @lucasheriques (+1000000 great stuff -adam)
• Adam's new UI for product tours is so good @lucasheriques (ty!!!! -adam)
• lots of great feedback on tours so far; looking at shifting alpha->beta once we get some reliably running internally

Retro - Tasks (WIP)

• 🟢 Collect product tours alpha feedback & iterate @adboio
• 🟢 Cross-product in-app experience RFC @adboio - rfc written https://github.com/PostHog/requests-for-comments-internal/pull/968, first-pass PR up for interim feat: coordinate in-app product interactions, first pass posthog-js#3033
• Surveys Everywhere work @adboio
  • 🟢 Surveys x LLM Analytics - v0 in PR feat(surveys): integrate with LLMA #46769, shipping soon!
  • 🔴 Reviewing other cross-sell data & iterating - cross-sell focus has been on llma
• ⚪ Surveys Performance @lucasheriques: talk with: Lucas R and Robbie to see what we can we do for a more performant cache for surveys
• 🔴 Surveys Platform UX @lucasheriques : simplify the screen for a survey that is created but not launched

Side Quest

• 🔴 see if we can improve the selector for actions to support tailwind classes with brackets (or if there's a better way to make this selector) @adboio
• 🟢 how hard is it to build native embed plugins for sites like framer/webflow? support ticket ref, i think we should make this easier for surveys. it's currently impossible haha @adboio
  • embeddable surveys have arrived! https://drive.google.com/file/d/1rm4nWDr97TTjO4EES7-IZwGnbdud2b6D/view?usp=drive_link
• ⚪ Allow surveys to be non dismissable on a question-basis @lucasheriques
• ⚪ Setup mobile environment to test out Android and iOS changes. Got a demand for surveys customizations from a big customer, wanna see if I can get to it next sprint @lucasheriques (@adboio did a lot of mobile work tohugh)

Quarterly goals

• Goal 1: Survey Platform UX Simplification @lucasheriques
• Goal 2: Product Tours Shipped @adboio
• Goal 3: Surveys Everywhere v2 @lucasheriques @adboio @cory
• Goal 4: PostHog AI User Interviews @lucasheriques @adboio @cory

This Sprint

High priority - WIP

• @lucasheriques ship new survey details page, with easier actions / easier AI insights
• @lucasheriques RFC on what does AI surveys look like? i.e. ask a follow up question on a survey
• 🟡 @adboio ‼️ ship first-pass at surveys-tours-convos coordination feat: coordinate in-app product interactions, first pass posthog-js#3033
• 🟡 @adboio scheduling lots of interviews with tours alpha users - what's missing or broken? what sucks? what's preventing beta launch?
• 🟡 @adboio ship first version of surveys<>llma, talk with customers about it
• @fivestarspicy look at mobile surveys usage

[hpouillot]

Team Error Tracking

Retro 🟢 🔴

@ablaszkiewicz enable redis encryption 🟢
@ablaszkiewicz pick up ingestion controls from david 🟢
@ablaszkiewicz stretch: spike alert configuration 🔴

@hpouillot cymbal refactoring 🟢
@hpouillot dsym support help mobile team 🔴
@hpouillot stretch: esbuild plugin 🔴

@fivestarspicy create dashboards for ET (LLMA insights, NPS Score) ✋

OKR

Alerting - Spikes & Digest
Engagement - Search & Speed
Grouping - Fingerprint & Sourcemaps
Collection - Cymbal refactoring & Mobile support

High priority

@ablaszkiewicz spike alert configuration (UI + cymbal imports)
@ablaszkiewicz wrong implementation fight with LLM (docs, warnings...)
@ablaszkiewicz stretch: feedback button

@hpouillot adapting similar issues to new embedding tables
@hpouillot fingerprint algorithm (source context)
@hpouillot fix search bar

@fivestarspicy metrics / dashboard
@fivestarspicy alerts product tour
@fivestarspicy logs integration POV

[robbie-c]

Team Analytics Platform & Customer Analytics

Support hero: @vdekrijger
Off during the sprint: @arthurdedeus from 13th

Quarter goals

• Do everything in the Query Performance mega-issue
• Iterate on schema validation
• 99.9% uptime for alerts, subscriptions, exports, and query service
• Improve query cache stability
• Get on top of support / open github issues
• Re-imagine dashboards
• Do a hackathon around https://github.com/PostHog/requests-for-comments-internal/pull/941
• Terraform provider beta hardening / path to GA

Retro

Tasks

@vdekrijger

• 🟢 Make sure all export failures are classified
• 🟢 Finish CI improvements for CH compatibility
• 🟢 When subscribing to dashboard, let people choose which insights are included

@andyzzhao

• 🟢 Finish up export fix to support buffering to disk for all query types
• Adding cursor pagination rather than offset pagination to some query runners
• Query cache on its own elasticache instance
  • Not sure if needed any more, for now going to focus elsewhere
    Extra: 🟢 pre-sort events query

@aspicer

• 🟡 Ship EAV using a replicated table for now
• 🟢 Get ingestion's review on EAV PR
• 🟢 Query schema ingestion blocking
  Extra: 🟢 lazy join bugs in hogql

@robbie-c

• 🟢 Look at creating a cohort for every team for their internal test filters
• 🔴 Kill off non-PoE queries

@arthurdedeus

• Revenue data in customer profiles #44006
• fix(customer-analytics): implement PowerUsersQueryRunner to fix timeout issues #45745
• Global customer journeys #44007
• Make customer profile nodes tab aware #45359
• Option to filter out test users from power users table #45910

Plan

High priority

@vdekrijger

• Release full RBAC control through TF and support customer during adoption
• Add quick filters to dashboards / insights
• Support hero work:
  • Want to tackle papercuts during my "downtime" as support hero
  • Look into reducing the high "alert delivery failed %" (around ~10% still)

@andyzzhao

• See if we still need cursor pagination
• Look into surprisingly low cache hit rate

@aspicer

• Schema ingestion validations (e.g. string literals, other value checks)
• Workarounds for super properties in type checking

@robbie-c

• Kill off non-PoE queries
• Working with experiments team to reuse exposures subquery between queries

@arthurdedeus

• Global customer journeys #44007
• Plan some customer interviews

Low priority / side quests

@arthurdedeus

• get ingestion team's thoughts on a Last Seen (truncated to e.g. hour) property

Are any other teams impacted by this plan? If so, tag them here

Blocked by CH:

• See https://posthog.slack.com/archives/C076R4753Q8/p1770224418225679?thread_ts=1768426673.029629&cid=C076R4753Q8

[rafaeelaudibert]

Team Growth

Support hero: @rafaeelaudibert and @MattBro
Off during the sprint: @MattBro on 9th-16th, @rafaeelaudibert on Tuesday 10th

Quarter goals

1. integrations with external partners 🟡
2. Improve sidebar cross-selling algorithm ⚪
3. working on improving the post-onboarding experience by making it less overwhelming 🟡
4. running tiny experiments in search of huge ROI ⚪

Details

⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

Retro

• Launch new quick start (@rafaeelaudibert) ✅
• Launch new onboarding/resource creation for at least a couple products (@rafaeelaudibert) 👷🏻
• Enabled managed_viewsets from Revenue Analytics for everyone and remove old code (@rafaeelaudibert) 🚫
• Iron out MCP OAuth bugs (@MattBro) ✅
• Push ReplIt on the connector (@MattBro) 👷🏻
• Go through the list and get everyone to use our MCP in their vibecoding platforms (@MattBro) 🚫
  • Also, see if they would be open to using PostHog as a platform for analytics, error tracking, session replay, and feature flags

Low priority / side quests

• Release sampo on posthog-elixir (@rafaeelaudibert) ✅
• Make sure MCP is working on V0 and Lovable (@MattBro) ✅

What else did we do?

• Vercel integration finished on our end, just waiting on their launch @MattBro
• Onboarded @fercgomes! @rafaeelaudibert
• Almost finished AI onboarding @MattBro

Plan

High priority

• Plan Fernando's quarter (@rafaeelaudibert)
• Keep improving onboarding state (specially Session Replay) (@rafaeelaudibert)
• Enabled managed_viewsets from Revenue Analytics for everyone and remove old code (@rafaeelaudibert)
• Support Vercel launch (@MattBro, @rafaeelaudibert while Brooker is out)
• Go through the list and get everyone to use our MCP in their vibecoding platforms (@MattBro)
• Chug through the onboarding list, work on the surveys for Workflows (@fercgomes)
• OAuth Authentication in the Toolbar (@fercgomes)
• Book the Boston and Barbados trip (@fercgomes)

Low priority / side quests

• Move posthog-php and posthog-python to sampo (@rafaeelaudibert)
• Posthog Dashboard on MCP usage (@MattBro)
• Investigate what we could change on the Growth dashboard (@fercgomes)

[kappa90]

Team PostHog AI

Support hero: @skoob13 @cvolzer3
Off during the sprint: none

Quarter goals

1. Quarter goals, status (red/yellow/green) and action points if yellow/red

Retro

🟠🟢🔴
🟠 @kappa90 new MCP for Twig (with write tools)
🔴 @skoob13 file system exploration (dumping, read syncing, checkpointing)
🔴 @kappa90 PostHog AI-inspired prompt injections in the Twig monorepo
🔴 @tatoalo use case skills in the Twig monorepo (wizard (set up events, errors, logs), feature flags -> "Launch a new feature" / "Setup up an A/B test", "How is my business going")
🟢 @kappa90 @tatoalo create feature flag/experiment/survey -> converted with clean schemas
🟠 @cvolzer3 convert MaxTools (revenue/web)

Plan

High priority

• @tatoalo new infra
  • kafka stream
  • ingestion service (capture logs agent)
  • agent logs node on CH in logs cluster
  • CH migrations in monorepo
  • twig integration
• @joshsny @kappa90 agent server that runs in the sandbox
  • SDK harness
  • ACP JSON stream protocol
  • websocket connection
• @skoob13 MCP 2.0 with SQL
  • Claude Plugin with skills
  • MCP codegen
  • feedback loops with customers (e.g. talk 10 times to PHAI -> email)
• @cvolzer3 MCP store (remote first)
• @kappa90 deliver MCP Apps to Twig folks

Low priority / side quests

• @kappa90 check skills state with docs folks
• @joshsny Slack interface? cc @Twixes

Are any other teams impacted by this plan? If so, tag them here

• @PostHog/team-docs-wizard

[adamleithp]

Team Plaform UX'cellence

Off during the sprint: N/A

Legend

⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned 🟣 = Ongoing

Q1 2026 Quarter goals We have diverged...

1. "Posthog is confusing" - reduce visual complexity of posthog and try to ease the label "confusing" or "space ship dashboard"
2. Twig ?
3. Continue to own platform UI/UX
   • Comb through platform and fix issues/spark joy wherever possible

Retro

High priority

1. 🟣 "posthog is confusing" - just need to release it!
   • 🟢 Remove sidepanel buttons (visually)
   • 🟢 New navigation menus to support moved items from side panel
     • 🟢 I moved the side panel to render "in the scene"
   • 🟢 Move context aware items into scene panel
     • 🟢 I moved the side panel to render "in the scene"
     • 🟢 AI, notebooks, access control and support still function as "sidepanel" items, but now live behind a ... item

Low priority / side quests

1. 🟣 Design system
   • 🔴 New color system (Shadcn-style)
   • 🟣 New UI library (Base_ui) - migration underway
   • 🔴 Add kick-ass documentation
2. 🟢 Released new search cmd k (modal based / new tab), hackable, quick and sparks joy
3. 🟢 Fixed issues/papercuts for customers
4. 🟣 Memory leak problem
5. 🟢 Health page (basis for new pages + @PostHog/team-growth work)
6. 🟢 PostHog status indicator + drill in cmd shift p
7. 🟢 Released new tooltip system - these are exactly the same, but when hovering a second item with a tooltip there's no delay, which really makes a big difference in exploration and ease of use.

Plan

High priority

1. 🟡 "posthog is confusing"
   • ⚪ Create offramp for changes (teach users where stuff is)
   • 🟢 Release side panel removal flag to team
   • ⚪ Release side panel removal flag to general public
     • ⚪ Handle feedback
     • ⚪ Coordinate docs update
2. 🟡 Ai first changes
   • Nav item "Posthog AI" which goes to fullscreen /ai with chat history panel
     • 🟢 Release side panel removal flag to team
     • ⚪ Release side panel removal flag to general public
     • 🟡 Handle feedback
     • ⚪ Coordinate docs update
   • Integrate home page with AI
     • 🟢 Release side panel removal flag to team
     • ⚪ Release side panel removal flag to general public
     • ⚪ Handle feedback
     • ⚪ Coordinate docs update
3. 🟣 Continue to own platform UI/UX
   • Comb through platform and fix issues/spark joy wherever possible

Low priority / side quests

3. ⚪ Twigv