fix(signals): deliver note-only report edits to Slack as the note

[andrewm4894]

Problem

• A team with a scout Slack destination sees the same report message twice when a scout later adds a note to a report it already posted.
• edit_report enqueues the same delivery as a fresh emit, so the worker re-renders the unchanged title and summary.
• The note, the only new content, never reaches the channel.

Changes

• A note-only edit now posts the note itself: a context line saying the scout added a note to an existing report, the report title, the note body, and the usual "View report in PostHog" button.
• _do_edit_report marks a delivery as note-only when a note was appended and the title and summary were untouched. An edit that rewrites content still posts the report message, even when it also appends a note.
• The note snapshot rides the Celery payload, clipped to 6,000 chars. Snapshotting at enqueue keeps the message on the note this edit added, not whatever the work log holds at send time.
• enqueue_scout_slack_delivery omits the edit_note kwarg when unset, so deliveries without a note keep the payload shape workers on the previous task signature accept during a rolling deploy.

What the channel sees when a scout adds a note to a report it already posted (mockups, not screenshots):

Before, a second copy of the report message:

Scout · Support trends
Login failures spiked after the 1.2 release
(the full report summary again, identical to the message above it)
View report in PostHog

After, the note framed as an update:

Scout · Support trends added a note to an existing report
Login failures spiked after the 1.2 release
Re-checked this morning: error rate is back to baseline since the fix deployed.
View report in PostHog

The note body gets the same treatment as a summary: markdown to mrkdwn conversion, mention-injection escaping, chart references reduced to labels, section truncation.

No app UI changes. I could not post a rendered message to a Slack workspace from this session, so the examples above are mockups.

How did you test this code?

• New test in test_scout_slack_delivery.py: a delivery carrying a note renders the note, not the summary. Catches a revert of the note branch, which would re-post the duplicate report message.
• Extended test_report_emit_and_edit_enqueue_configured_slack_destination_after_commit: only the note-only edit carries edit_note; an emit and a summary rewrite with a note do not. Catches the discriminator regressing to "any note".
• New test: the enqueue omits the edit_note kwarg when unset. Catches a payload-shape change that would make old workers reject every delivery during a rolling deploy.
• Not run to completion locally: the Django suite needs a database this sandbox had to bootstrap from scratch, and the run was still migrating at PR time. CI runs these files.

Automatic notifications

• Publish to changelog?

Docs update

None: the scout Slack destination behavior is not separately documented.

🤖 Agent context

Autonomy: Human-driven (agent-assisted)

• Written by Claude Code in a remote session directed by a PostHog engineer; the driver asked for note-only report edits to stop reading as duplicate Slack messages.
• Skills invoked: /writing-user-facing-copy, /writing-tests, /writing-code-comments, /writing-pr-descriptions.
• Considered re-reading the note from the report's work log at send time instead of snapshotting it into the payload, and rejected it: later notes could swap in, and the delivery task cannot tell which note this delivery was for.

🤖 Generated with Claude Code

https://claude.ai/code/session_01DYWXRYPWd3cK7pza5gjHd4

[trunk-io]

😎 Merged successfully - details.

[posthog]

🦔 ReviewHog reviewed this pull request

Found 0 must fix, 1 should fix, 0 consider.

Published 1 finding (view the review).

[greptile-apps]

_{Reviews (1): Last reviewed commit: "fix(signals): deliver note-only report e..." | Re-trigger Greptile}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 80f50d16dc

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[posthog]

ReviewHog Alpha 🦔 If you find any issues helpful - please reply "valid", "invalid", etc., for evaluation purposes 🙏

[posthog]

ReviewHog Report

Found 1 should fix.

[trunk-io]

   

View Full Report ↗︎ ⋅ Docs

[andrewm4894]

/trunk merge

[stamphog]

Note

🤖 stamphog reviewed 24d52e46bbf8721b46a1f7a3641bafbf33b16e2a — verdict: ESCALATE

A ReviewHog finding confirms this PR's producer can send an edit_note kwarg to a Celery worker still running the previous task signature during a rolling deploy, dropping the note notification with a TypeError. The bot explicitly said this is a release-sequencing tradeoff for a human to decide, not something it can resolve, and the thread remains unresolved with no human sign-off on accepting that risk.

• Author wrote 0% of the modified lines and has 143 merged PRs in these paths (familiarity MODERATE).
• 👍 on the PR from greptile-apps[bot].
• Unresolved inline thread on products/signals/backend/tasks.py: note-only Slack deliveries enqueued during the rollout of this PR can land on an old worker that rejects the new edit_note kwarg, dropping the note notification (bounded to this one deploy, report data itself is unaffected).
• Author (via Claude Code) argued for accepting this risk rather than fixing it, and offered a two-stage deploy alternative on request, but no maintainer has weighed in to accept that tradeoff.
• No approving review from a human or an agent reviewer with no outstanding concerns is present — all reviews are COMMENTED.

Gate mechanics and policy version

Gate		Result
prerequisites	✓	all clear
deny-list	✓	no deny categories matched
size	✓	99L, 4F substantive, 182L/7F incl. docs/generated/snapshots — within ceiling
tier	✓	T1-agent / T1c-medium (182L, 7F, single-area, fix)
stamphog 2.0.0b4		.stamphog/policy.yml @ aa7b182 · reviewed head 24d52e4

[deployment-status-posthog]

Deploy status

Environment	Status	Deployed At	Workflow
dev	✅ Deployed	2026-08-14 18:06 UTC	Run
prod-us	✅ Deployed	2026-08-14 18:28 UTC	Run
prod-eu	✅ Deployed	2026-08-14 18:30 UTC	Run