[posthog]

Problem

• New people keep the Recommended inbox order after their first activity summary lands, until a personal ranking tick builds their list.
• Follows #133442, which made the first summary arrive sooner.

Closes #133427

Origin

• Team member
• First signal: 2026-11-11
• Inbox report: open
• Task started by: a person, from the inbox

Changes

• A person's first list now builds as soon as their first summary is stored.
• The build uses the on-demand gates and runs only when it calls the model. Otherwise the tick keeps the pair.
• A new first_summary trigger marks these builds.
• The hook is in the tick's run_pair, because a call from update_activity_summary creates an import cycle.

The activity summary tick now also starts a personal ranking build:

flowchart LR
  A[Activity summary tick]:::y --> B[Store summary]:::g
  C[Personal ranking tick]:::y --> D[Build list]:::b
  classDef y fill:#f9bd2b,stroke:#f9bd2b,color:#000;
  classDef g fill:#e5e7eb,stroke:#c7ccd1,color:#000;
  classDef b fill:#1d4aff,stroke:#1d4aff,color:#fff;

Loading

flowchart LR
  A[Activity summary tick]:::y --> B[Store summary]:::g
  B -- first summary, no list --> E[Celery build, first_summary]:::b
  C[Personal ranking tick]:::y --> D[Build list]:::b
  classDef y fill:#f9bd2b,stroke:#f9bd2b,color:#000;
  classDef g fill:#e5e7eb,stroke:#c7ccd1,color:#000;
  classDef b fill:#1d4aff,stroke:#1d4aff,color:#fff;

Loading

How did you test this code?

Test rationale:

• A parameterized case covers candidates, an existing list, a disabled payload, and no candidates.
• The OK-update tick test now asserts that the build is queued.
• Not run: a real Celery worker.

Release status

• This change is behind a feature flag and is not available to users

Docs update

• Updated products/signals/backend/management/AGENTS.md.

🤖 Agent context

Autonomy: Human-driven (agent-assisted)

Agent: Claude Code, Opus 5.5 (claude-opus-5-5)

Skills: /writing-tests, /writing-code-comments, /editing-agents-md, /writing-pr-descriptions

More context for AI agents

• Session: PostHog Desktop task from an inbox report. Tools: PostHog MCP, pytest via hogli.
• Decisions: The build skips tick enrollment, as the report asked. Cost stays bounded by the shared limit and the candidates requirement.
• Decisions: A second or later summary queues nothing. It reaches the tick as changed_activity_summary_version.
• Read closely: personal_ranking_tasks.py, start_first_summary_build and _skip_build.
• Follow-ups: Within 48 hours of deploy, check for personal inbox ranking built events with trigger = first_summary.

---

Created with PostHog Desktop from an inbox report

🤖 Generated with Claude Code

[trunk-io]

😎 Merged successfully - details.

[parameterai]

Risk: No findings

This PR queues a build_personal_inbox_ranking Celery task (new first_summary trigger) from the activity summary tick when a person's first summary is stored and they have no ranking row. I traced the new path end to end: the build is gated by the flag's enabled, the per-person cache lock, no-existing-row, model-only execution (which folds in AI data-processing approval via may_call_model), and the shared max_builds_per_day bucket, so I found no exploitable bypass.

_{Sentinel reviewed f0de76b · Review settings}

[github-actions]

🤖 CI report

⚠️ Trunk lane — backend Python lane

This PR is assigned to the backend Python lane. It runs backend Python tests and may merge in parallel with PRs in other lanes.

✅ Duplication (Python) — clean

New Python code duplication introduced by this branch. Fails at 70+ tokens in app code, or 150+ tokens when both copies live in test files. Advisory while the gate proves itself: extract a shared helper instead of copying.

✅ Duplication (TypeScript) — clean

New TypeScript code duplication introduced by this branch. Fails at 70+ tokens in app code, or 150+ tokens when both copies live in test files. Advisory while the gate proves itself: extract a shared helper instead of copying.

⚠️ Backend coverage — 89.0% of changed backend lines covered — 6 uncovered

🧪 Backend test coverage

Patch coverage — changed backend lines (products + core): ██████████████████░░ 89.0% (49 / 55)

File	Patch	Uncovered changed lines
products/signals/backend/activity_summary/tick.py	66.7%	411, 413
products/signals/backend/personal_ranking_tasks.py	81.8%	79, 82–84

🤖 Agents: add a test only if an uncovered line exposes a realistic regression that existing tests miss. Otherwise explain why no new test is needed under "How did you test this code?". Gap list: the patch-coverage artifact on this run (gh run download 321291394039840 -n patch-coverage), or the coverage-data block at the end of this comment.

Per-product line coverage (touched products)

Product	Coverage	Lines
platform_features	██░░░░░░░░░░░░░░░░░░ 12.1%	7 / 58
data_tools	████████████░░░░░░░░ 57.6%	95 / 165
warehouse_sources_queue	█████████████░░░░░░░ 65.8%	2,192 / 3,330
demo	██████████████░░░░░░ 71.3%	1,978 / 2,775
knowledge_writer	████████████████░░░░ 77.7%	1,409 / 1,813
batch_exports	█████████████████░░░ 82.9%	24,782 / 29,885
engineering_analytics	█████████████████░░░ 85.6%	12,515 / 14,615
ml_inference	█████████████████░░░ 86.3%	524 / 607
commands	█████████████████░░░ 86.5%	1,645 / 1,901
mcp_analytics	██████████████████░░ 87.8%	6,827 / 7,777
posthog_ai	██████████████████░░ 88.2%	4,514 / 5,116
apm	██████████████████░░ 88.2%	1,973 / 2,236
notebooks	██████████████████░░ 88.6%	15,594 / 17,599
dashboards	██████████████████░░ 89.5%	8,969 / 10,017
cohorts	██████████████████░░ 90.7%	9,866 / 10,883
warehouse_suggestions	██████████████████░░ 90.7%	4,264 / 4,700
actions	██████████████████░░ 90.9%	1,524 / 1,677
visual_review	██████████████████░░ 90.9%	10,634 / 11,698
data_warehouse	██████████████████░░ 91.1%	15,984 / 17,546
signals	██████████████████░░ 91.2%	112,499 / 123,339
cdp	██████████████████░░ 91.4%	10,845 / 11,865
today	██████████████████░░ 91.5%	4,136 / 4,521
canvas	██████████████████░░ 91.5%	8,559 / 9,351
tasks	██████████████████░░ 91.6%	86,251 / 94,204
autoresearch	██████████████████░░ 91.6%	12,203 / 13,318
data_modeling	██████████████████░░ 91.7%	12,686 / 13,831
streamlit_apps	██████████████████░░ 92.0%	3,200 / 3,478
alerts_platform	██████████████████░░ 92.0%	10,654 / 11,576
release	██████████████████░░ 92.1%	453 / 492
metrics	██████████████████░░ 92.1%	12,068 / 13,102
business_knowledge	██████████████████░░ 92.2%	10,900 / 11,824
review_hog	██████████████████░░ 92.3%	18,397 / 19,930
conversations	██████████████████░░ 92.4%	31,728 / 34,335
wizard	██████████████████░░ 92.4%	7,323 / 7,922
error_tracking	███████████████████░ 92.5%	20,976 / 22,675
slack_app	███████████████████░ 92.6%	25,544 / 27,600
early_access_features	███████████████████░ 92.7%	1,377 / 1,486
notifications	███████████████████░ 92.8%	1,230 / 1,326
legal_documents	███████████████████░ 93.0%	2,380 / 2,558
ai_training	███████████████████░ 93.1%	605 / 650
managed_warehouse	███████████████████░ 93.2%	15,792 / 16,947
authorization	███████████████████░ 93.2%	850 / 912
alerts	███████████████████░ 93.2%	13,753 / 14,756
webmcp	███████████████████░ 93.3%	277 / 297
stamphog	███████████████████░ 93.3%	11,441 / 12,265
web_analytics	███████████████████░ 93.5%	27,059 / 28,934
surveys	███████████████████░ 93.6%	7,847 / 8,388
exports	███████████████████░ 93.7%	4,474 / 4,777
approvals	███████████████████░ 93.7%	5,020 / 5,359
cross_project_dashboards	███████████████████░ 93.7%	867 / 925
context_layer	███████████████████░ 93.9%	4,075 / 4,342
messaging	███████████████████░ 93.9%	9,228 / 9,827
marketing_analytics	███████████████████░ 93.9%	22,429 / 23,882
mcp_store	███████████████████░ 94.0%	9,391 / 9,991
ai_gateway	███████████████████░ 94.0%	1,472 / 1,566
workflows	███████████████████░ 94.0%	29,658 / 31,549
billing_alerts	███████████████████░ 94.0%	2,332 / 2,480
annotations	███████████████████░ 94.1%	1,460 / 1,552
managed_migrations	███████████████████░ 94.2%	1,886 / 2,003
customer_analytics	███████████████████░ 94.2%	31,699 / 33,658
ai_observability	███████████████████░ 94.3%	38,322 / 40,635
access_control	███████████████████░ 94.3%	8,606 / 9,122
replay_vision	███████████████████░ 94.4%	33,431 / 35,422
logs	███████████████████░ 94.4%	20,533 / 21,748
reminders	███████████████████░ 94.5%	837 / 886
experiments	███████████████████░ 95.0%	47,368 / 49,882
warehouse_sources	███████████████████░ 95.0%	142,320 / 149,854
growth	███████████████████░ 95.0%	11,602 / 12,216
endpoints	███████████████████░ 95.0%	9,725 / 10,239
product_tours	███████████████████░ 95.1%	1,271 / 1,337
skills	███████████████████░ 95.1%	7,658 / 8,053
uptime_monitoring	███████████████████░ 95.3%	306 / 321
data_catalog	███████████████████░ 95.6%	4,749 / 4,965
tracing	███████████████████░ 95.7%	5,087 / 5,315
ui_customization	███████████████████░ 95.7%	714 / 746
subscriptions	███████████████████░ 95.7%	14,176 / 14,811
product_analytics	███████████████████░ 95.8%	31,450 / 32,836
feature_flags	███████████████████░ 95.8%	30,752 / 32,104
revenue_analytics	███████████████████░ 95.9%	1,877 / 1,958
aeo	███████████████████░ 95.9%	4,131 / 4,308
data_quality	███████████████████░ 96.0%	8,583 / 8,937
groups	███████████████████░ 96.4%	2,068 / 2,145
account_audits	███████████████████░ 96.5%	823 / 853
security	███████████████████░ 96.8%	1,597 / 1,650
pulse	███████████████████░ 97.0%	2,026 / 2,089
analytics_platform	████████████████████ 98.0%	2,962 / 3,022
field_notes	████████████████████ 98.4%	181 / 184

Report-only. Patch coverage = changed backend lines covered vs origin/master. Sorted lowest first.
Known gaps: lines covered only by Temporal tests show as uncovered; core line numbers may drift if master changed the same file.

[andrewm4894]

/trunk merge

[deployment-status-posthog]

Deploy status

Environment	Status	Deployed At	Workflow
dev	✅ Deployed	2026-11-11 16:39 UTC	Run
prod-us	✅ Deployed	2026-11-11 18:14 UTC	Run
prod-eu	✅ Deployed	2026-11-11 17:23 UTC	Run

_{This table shows the posthog release only. Hogstation shows every service that this PR reaches, Kargo services included.}