newsletters

name	newsletters
description	Full newsletter lifecycle — create, configure all settings (frequency, personalization, AI model, web search, reply threading, plan mode wiring), manage subscribers, attach forms, draft and manage issues, preview personalized letters, enable private chat and email replies, and test in dev.
argument-hint	[create | settings | subscribers | forms | drafts | replies | dev]

Newsletters — Complete Reference

Newsletters are AI-powered publications scoped to a datasphere. Every setting is configurable: schedule, personalization mode, AI model, web search, reply threading, and planner task wiring.

---

The Personalized Newsletter Pattern

Subscriber fills out a survey (the "form")
        ↓
Their answers are stored as subscriber metadata
        ↓
At send time: AI reads their metadata + searches the web for timely context
        ↓
AI generates a unique piece of content written specifically for them
        ↓
They receive something that feels like it was made just for them
        ↓
Every letter includes a link to update their profile anytime

---

1. Create a Newsletter

create_newsletter(
    datasphereId="...",                  # DB ID — get from list_dataspheres
    name="Love You From You",
    description="Weekly self-love letters, written just for you",
    personalizationMode="PERSONALIZED",  # NONE | TEMPLATE | PERSONALIZED
    systemInstructions="...",            # Editorial brief — see templates below
    globalContext="...",                 # Persistent facts AI reads every run
    enableWebSearch=True,
    webSearchInstructions="find 2-3 self-care practices relevant to {current_struggle}",
    aiModel="claude-sonnet-4-6",         # claude-sonnet-4-6 | claude-opus-4-7 | claude-haiku-4-5-20251001
    scheduleType="WEEKLY",               # MANUAL | DAILY | WEEKLY | BIWEEKLY | MONTHLY | CUSTOM
)

HIL flow: gather the 5 dimensions first (voice, personalization inputs, timely context, structure, subscriber journey), write systemInstructions in your response and show it to the user before calling.

systemInstructions Templates by Voice Type

First-person FROM subscriber (self-love, journaling, affirmations):

"You are writing a [description] in the first person, as if the subscriber is writing to themselves at their wisest and most compassionate. Use only 'I', 'me', 'my'. Never 'you' or 'your'. Draw entirely from their survey answers — every sentence should feel like it could only have been written for this specific person. [Structure]. End with [closing element]."

Second-person TO subscriber (coaching, mentorship, accountability):

"You are a [role description] writing directly to the subscriber. Use warm second-person ('you', 'your'). Reference their specific answers throughout — never generic advice. [Structure]. Tone: [adjectives]."

Third-person narrator (astrology, bedtime stories, horoscopes):

"You are writing a personalized [format] for the subscriber. Use their birth data / profile answers to make every insight specific to them, not just their sun sign. Incorporate web search results as current cosmic/contextual context. [Structure]."

---

2. Update All Newsletter Settings

Use update_newsletter to change any setting after creation. Show a before/after preview before calling.

update_newsletter(
    newsletterId="nl_...",

    # Identity
    name="New Name",
    description="New tagline",

    # AI editorial brief
    systemInstructions="...",     # Most important — full editorial brief
    globalContext="...",          # Persistent facts: names, dates, locations

    # AI model
    aiModel="claude-opus-4-7",    # Upgrade for richer letters
    contextWindowDays=30,         # Days of datasphere activity AI looks back

    # Personalization
    personalizationMode="PERSONALIZED",   # NONE | TEMPLATE | PERSONALIZED
    contentTemplate="...",                # TEMPLATE mode only — use {{firstName}}, {{metadata.FIELD}}

    # Web search
    enableWebSearch=True,
    webSearchInstructions="find this week's {sun_sign} horoscope and planetary transits",

    # Schedule
    scheduleType="WEEKLY",        # MANUAL | DAILY | WEEKLY | BIWEEKLY | MONTHLY | CUSTOM

    # Planner task wiring
    planModeId="...",             # Board ID — ALL tasks in this board flow into AI context
    taskStatusFilter=[],          # [] = all statuses (default). ["TODO","IN_PROGRESS"] = open only

    # Reply threading
    private_thread_enabled=True,  # Enable subscriber email reply chat
    billing_mode="DATASPHERE_POOL",  # SUBSCRIBER | DATASPHERE_POOL
)

Settings Reference

Setting	Values	Effect
name	string	Display name shown in subscriber emails and the admin UI
description	string	Internal tagline — not shown to subscribers
systemInstructions	string	Full editorial brief the AI reads at generation time — the most important field
globalContext	string	Persistent facts (names, dates, locations) prepended to every AI context window
personalizationMode	NONE	Same content for every subscriber
	TEMPLATE	HTML with {{firstName}}, {{metadata.FIELD}} variables
	PERSONALIZED	AI generates a unique letter per subscriber using their survey answers
scheduleType	MANUAL	Admin triggers each issue manually
	DAILY / WEEKLY / BIWEEKLY / MONTHLY	Automated cadence
	CUSTOM	Custom cron
aiModel	claude-haiku-4-5-20251001	Fast, lightweight — good for simple updates
	claude-sonnet-4-6	Default — best balance
	claude-opus-4-7	Richest output — use for high-stakes personalized letters
contextWindowDays	number	How many days of datasphere pages/tasks the AI reads
contentTemplate	HTML string	TEMPLATE mode only — use {{firstName}}, {{metadata.FIELD}} variables
enableWebSearch	true / false	Let AI search web at generation time
webSearchInstructions	string	Search query template with {variable} slots — e.g. "find this week's {sun_sign} horoscope"
planModeId	board ID	Wire a planner board — all tasks become AI context (great for event calendars)
taskStatusFilter	[]	All tasks. ["TODO","IN_PROGRESS"] = open only
private_thread_enabled	true	Issued emails include a unique reply-to per subscriber → group chat
billing_mode	DATASPHERE_POOL	AI reply costs come from the datasphere capacity pool
	SUBSCRIBER	Each subscriber pays from their own capacity

Planner wiring pattern: Create a board (e.g. "Astrology Events"), add events as tasks with due dates, set planModeId. Every task in that board flows into every issue's AI context automatically — no extra steps.

---

3. Subscriber Management

Add a subscriber

add_newsletter_subscriber(
    newsletterId="nl_...",
    email="[e-mail]",
    name="Jane Smith",
    phoneNumber="+1...",       # optional — SMS delivery
    userId="...",              # optional — link to platform user instead of email
    invite=False,              # False = ACTIVE immediately. True = send confirmation email (PENDING)
)

Two ways people subscribe:

1. Self-subscribe (recommended for public) — share /newsletters/{slug}. Visitors see the archive and subscribe form.
2. Manual add (this tool) — for invited guests, beta testers, known contacts.

List subscribers

list_newsletter_subscribers(
    newsletterId="nl_...",
    status="ACTIVE",  # optional filter: ACTIVE | PENDING | UNSUBSCRIBED
)

Returns: id, email, name, status, metadata, surveyCompletionStatus, createdAt.

Status meanings:

• ACTIVE — receiving emails
• PENDING — invite sent, awaiting email confirmation
• UNSUBSCRIBED — opted out (never re-add without consent)
• BOUNCED — delivery failed

Remove a subscriber

remove_newsletter_subscriber(
    newsletterId="nl_...",
    subscriberId="sub_..."  # get from list_newsletter_subscribers
)

Hard-delete — use only when explicitly requested. For unsubscribes, the subscriber portal handles that automatically.

Get a subscriber's profile (their personalization data)

get_subscriber_profile(
    subscriberId="sub_...",
    magicToken="..."  # optional — for anonymous subscriber access via magic link
)

Returns: every mapped survey question + their current answers + the metadata fields used by the AI. Use this to:

• Debug why a letter read as generic ("What does she know about herself so far?")
• Find which metadata fields are missing before generation
• Show a subscriber their own profile

Submit answers on behalf of a subscriber

submit_subscriber_answers(
    subscriberId="sub_...",
    answers=[
        {"questionId": "q_...", "answerFormat": "LONG_TEXT", "textAnswer": "I'm proud of..."},
        {"questionId": "q_...", "answerFormat": "MULTIPLE_CHOICE", "selectedChoices": ["choiceId1"]},
    ]
)

Answer formats: TEXT, LONG_TEXT, MULTIPLE_CHOICE, AUDIO (with audioFileUrl), VIDEO (with videoFileUrl). Idempotent — replaying the same payload overwrites previous answers.

Always call get_subscriber_profile first to know which questions exist and what's missing.

Get subscriber's issue history (journal)

get_subscriber_journal(subscriberId="sub_...")

Returns every issue sent to this subscriber + their personalized content. Useful for showing a subscriber their letter archive.

Bulk subscriber management (admin UI)

For CSV import/export and bulk actions, send the user to the admin page:

/app/{uri}/newsletters/{newsletterId}/subscribers

---

4. Forms — Create, Configure, Attach

Forms (surveys) are the personalization engine. Every answer a subscriber gives feeds into their metadata, which the AI uses to write their letter.

Create a survey

create_survey(
    title="Tell Me About You",
    datasphereUri="my-datasphere",
    surveyMode="QUESTIONNAIRE",
    surveyAccessLevel="PUBLIC",
    allowMultipleResponses=True,
    collectRespondentEmail=False,
    resultsAccessLevel="ADMIN_ONLY",
)
# → {"id": "page_...", "slug": "tell-me-about-you"}
# Save this pageId for all subsequent calls.

AI-generate questions

generate_survey_questions(
    pageId="page_...",
    prompt="This is a self-love newsletter. Generate 6-8 questions that give the AI enough to write a deeply personal letter. Include: first name, what they're proud of, current struggle, love language, affirmation they want to hear, preferred self-care, current energy level."
)

Review the generated questions, then add, edit, or reorder as needed.

Add a question manually

add_survey_question(
    pageId="page_...",
    questionText="What is your name?",
    answerFormats=["TEXT"],
    isRequired=True,
    order=0,
)

Answer formats:

Format	Use for	Needs choices?
TEXT	Short text	No
LONG_TEXT	Multi-line textarea	No
MULTIPLE_CHOICE	Single select	Yes
CHECKBOX	Multi-select	Yes — allowMultiple: true required
DROPDOWN	Single select dropdown	Yes
AUDIO_LIVE	Record audio in browser	No
AUDIO_UPLOAD	Upload audio file	No
VIDEO_LIVE	Record video	No
VIDEO_UPLOAD	Upload video file	No
IMAGE_UPLOAD	Upload image	No

CRITICAL for CHECKBOX: allowMultiple MUST be true. Choices format: { options: ["string1", "string2"] } — NOT objects.

Edit a question

update_survey_question(
    pageId="page_...",
    questionId="q_...",
    questionText="Updated question text",
    isRequired=False,
)

Delete a question

delete_survey_question(pageId="page_...", questionId="q_...")

Reorder questions

reorder_survey_questions(
    pageId="page_...",
    orderedIds=["q_001", "q_003", "q_002"]  # new order
)

List questions

list_survey_questions(pageId="page_...")

View responses

get_survey_responses(pageId="page_...")

Analytics

get_survey_analytics(pageId="page_...")
# → totalResponses, completionRate, distribution per question

Export responses

export_survey_responses(pageId="page_...", format="csv")  # csv | json

Link a form to a newsletter

This is the critical wiring step — connects the survey to the newsletter's AI context and generates personalized invite codes for every subscriber.

link_survey_to_newsletter(
    newsletterId="nl_...",
    surveyPageId="page_...",
    codePrefix="LOVE",   # appears in subscriber's invite URL
)

After this:

1. Survey appears with "In AI context" green badge in the survey editor
2. Per-subscriber personalized survey URLs are generated: /s/{uri}/{slug}?code={code}
3. Generation service automatically appends a "Update your profile" section + QR code to every sent letter
4. Subscribers who haven't responded get a "complete your profile" CTA instead of a personalized letter

Wire metadata fields (survey mappings)

After linking, tell the AI which question answer → which metadata field:

# Step 1 — let AI suggest all mappings at once
suggest_survey_mappings(
    newsletterId="nl_...",
    surveyPageId="page_...",
    prompt="This is a self-love newsletter. Map each question to a metadata field the letter AI can reference by name."
)
# Returns: [{questionId, metadataField, mappingType, reasoning}, ...]

# Step 2 — apply each suggested mapping
create_survey_mapping(
    newsletterId="nl_...",
    questionId="q_...",
    metadataField="current_struggle",   # snake_case — AI references this by name in letters
    mappingType="DIRECT",               # DIRECT | BOOLEAN | ARRAY_APPEND | OBJECT_MERGE | TRANSFORM
)

# Step 3 — verify
list_survey_mappings(newsletterId="nl_...")

Mapping types:

• DIRECT — copy answer as-is (most common)
• BOOLEAN — convert yes/no → true/false
• ARRAY_APPEND — add answer to a list (use for multi-select questions)
• OBJECT_MERGE — merge answer into an object
• TRANSFORM — apply custom rules

---

5. Drafting Issues (Safe — No Send)

Drafts sit in the queue with status=DRAFT. Nothing is sent until you explicitly call send_newsletter_issue.

Create a draft (write it yourself)

draft_newsletter_issue(
    newsletterId="nl_...",
    subject="Weekly Digest #3 — Leaning Into Rest",
    contentHtml="<h2>...</h2><p>...</p>",   # full email-safe HTML
    topicsCovered=["self-care", "rest", "boundaries"],
    contextSummary="Focus on rest after user mentioned burnout last week",
)

Recommended flow: Draft the issue content yourself in your response first. Show it to the user. Get edits. Then save with this tool.

Don't call this when the user is already on an issue editor page — use update_newsletter_issue instead to avoid creating a duplicate.

Update an existing draft

update_newsletter_issue(
    issueId="iss_...",
    subject="Updated subject line",
    contentHtml="...",     # APPEND or REPLACE — see below
    adminNotes="Private notes visible only to admins",
    topicsCovered=["updated", "topics"],
)

Append vs replace:

• "Add a section about X" → read current contentHtml, append a new <h3>+<p> block, send combined HTML
• "Rewrite the issue" → send only the new HTML
• "Change the subject" → send only subject, leave contentHtml unset

List all issues (drafts and sent)

list_newsletter_issues(newsletterId="nl_...")
# → [{id, subject, status (DRAFT|SENT), createdAt, sentAt, recipientCount}, ...]

AI-generate a draft

generate_newsletter_issue(
    newsletterId="nl_...",
    customPrompt="Focus on the theme of renewal and fresh starts this week",
)
# Takes 10-30 seconds. Returns subject + full contentHtml.
# For PERSONALIZED newsletters: this generates the base structure.
# Per-subscriber letters are generated at send time.

Always show the returned draft before asking if the user wants to send.

Send an issue (irreversible — real emails)

HIL gate: always show the draft and get explicit user confirmation before calling send.

send_newsletter_issue(issueId="iss_...")
# Sends to ALL ACTIVE subscribers. This is irreversible — emails are delivered immediately.
# For PERSONALIZED newsletters: one unique letter per subscriber generated at send time.

Email-safe HTML subset (only these tags render reliably across email clients):

Tag	Use
<h1>, <h2>, <h3>	Headings
<p>	Paragraphs
<strong>, <em>, <u>	Bold, italic, underline
<a href="...">	Links
<ul>, <ol>, <li>	Lists
<img src="..." alt="...">	Images (use hosted URLs)
<br>	Line break
<table>, <tr>, <td>	Tables (layout, not display: flex)
<blockquote>	Pull quotes

Do not use: <div>, <script>, <form>, CSS grid/flex, or inline style with shorthand properties.

---

6. Preview Personalized Letters

Before sending, always preview for a specific subscriber to verify personalization is wired:

preview_personalized_newsletter(
    newsletterId="nl_...",
    subscriberId="sub_...",   # pick one who has completed the survey
    customPrompt="...",       # optional — additional context for this preview only
)

Returns contentHtml (the actual letter they would receive) and a debug block with personalizationMode, metadataKeys, nameResolved.

If the letter reads as generic (same for everyone), check:

• personalizationMode is PERSONALIZED on the newsletter
• The subscriber has survey responses — call get_subscriber_profile
• link_survey_to_newsletter has been run (survey shows "In AI context" badge)
• list_survey_mappings shows all questions mapped to metadata fields

---

7. Reply Threading — Private Chat and Email Replies

Enable subscriber replies and group chat directly from newsletter emails:

update_newsletter(
    newsletterId="nl_...",
    private_thread_enabled=True,
    billing_mode="DATASPHERE_POOL",  # or SUBSCRIBER
)

How it works:

• Issued emails include a unique reply-to address per subscriber (unique Reply-To header)
• When a subscriber replies, it routes into a group chat thread
• Ari and other subscribers can respond in the thread
• Each reply is attributed to the correct subscriber

Billing modes:

• DATASPHERE_POOL — AI responses charged to the datasphere's community capacity pool. Any subscriber can participate regardless of tier. Requires an active paid capacity period on the datasphere.
• SUBSCRIBER — each subscriber is charged from their own capacity pool. Good for paid communities.

After enabling: "Issued emails will now include a reply footer. Subscribers reply to a unique address that routes into the group chat."

---

8. Verification Checklist

After full setup, verify:

• personalizationMode is PERSONALIZED on the newsletter
• systemInstructions is set (not empty)
• Survey shows "In AI context" green badge in the survey editor
• list_survey_mappings shows all questions mapped to metadata fields
• preview_personalized_newsletter returns a letter with debug.metadataKeys populated
• The preview letter reads as specific to that person, not generic
• At least one ACTIVE subscriber exists before sending

---

9. Dev / Local Testing

Use the local dev API key and local server for all testing before touching prod.

Local credentials (from ~/.dataspheres.env):

DAI_LOCAL_KEY=dsk_<REDACTED-set-in-~/.dataspheres.env>
Local server: http://localhost:3000

Test flow (PowerShell — Windows-native):

# 1. Point to local
$env:DATASPHERES_API_KEY = "dsk_<REDACTED-set-in-~/.dataspheres.env>"
$env:DATASPHERES_BASE_URL = "http://localhost:3000"

# 2. Create a test newsletter in a local datasphere
$body = @{name="Test Newsletter"; personalizationMode="PERSONALIZED"; scheduleType="MANUAL"} | ConvertTo-Json
$body | Out-File -Encoding utf8 "$env:TEMP\nl_create.json"
Invoke-RestMethod -Method Post -Uri "http://localhost:3000/api/dataspheres/{dsId}/newsletters" `
  -Headers @{"Authorization"="Bearer $env:DATASPHERES_API_KEY"} `
  -ContentType "application/json" -InFile "$env:TEMP\nl_create.json"

# 3. Create test subscriber
$sub = @{email="[e-mail]"; name="Test User"; invite=$false} | ConvertTo-Json
$sub | Out-File -Encoding utf8 "$env:TEMP\nl_sub.json"
Invoke-RestMethod -Method Post -Uri "http://localhost:3000/api/newsletters/{nlId}/subscribers" `
  -Headers @{"Authorization"="Bearer $env:DATASPHERES_API_KEY"} `
  -ContentType "application/json" -InFile "$env:TEMP\nl_sub.json"

# 4. Preview (no email sent)
$prev = @{subscriberId="{subId}"} | ConvertTo-Json
$prev | Out-File -Encoding utf8 "$env:TEMP\nl_preview.json"
Invoke-RestMethod -Method Post -Uri "http://localhost:3000/api/newsletters/{nlId}/preview" `
  -Headers @{"Authorization"="Bearer $env:DATASPHERES_API_KEY"} `
  -ContentType "application/json" -InFile "$env:TEMP\nl_preview.json"

On Windows, save JSON bodies to temp files:

$body = @{name="Test Newsletter"; personalizationMode="PERSONALIZED"} | ConvertTo-Json
$body | Out-File -Encoding utf8 "$env:TEMP\nl_body.json"
Invoke-RestMethod -Method Post -Uri "http://localhost:3000/api/dataspheres/{dsId}/newsletters" `
  -Headers @{"Authorization"="Bearer $env:DATASPHERES_API_KEY"} `
  -ContentType "application/json" `
  -InFile "$env:TEMP\nl_body.json"

E2E Playwright test:

npx playwright test tests/e2e/specs/newsletter-survey-personalized-draft-flow.spec.ts --workers=1

Dev vs Prod checklist:

• Verified full flow locally (create → subscriber → survey → link → mapping → preview) before promoting
• preview_personalized_newsletter tested with a real subscriber who has survey answers
• Reply threading tested with a test subscriber before enabling on prod
• send_newsletter_issue never called in local dev against real subscriber emails

---

API Reference

Tool	Method	Endpoint	Notes
list_newsletters	GET	/api/v2/dataspheres/:uri/newsletters	Uses URI (not DB ID)
create_newsletter	POST	/api/dataspheres/:datasphereId/newsletters	Uses DB ID
update_newsletter	PATCH	/api/newsletters/:newsletterId	All settings incl. reply threading
draft_newsletter_issue	POST	/api/newsletters/:newsletterId/issues	Creates DRAFT, no send
update_newsletter_issue	PATCH	/api/newsletter-issues/:issueId	Edit existing draft
list_newsletter_issues	GET	/api/newsletters/:newsletterId/issues	Drafts + sent
generate_newsletter_issue	POST	/api/newsletters/:newsletterId/generate	AI-generated draft
preview_personalized_newsletter	POST	/api/newsletters/:newsletterId/preview	No send — safe to call anytime
send_newsletter_issue	POST	/newsletter-issues/:issueId/send	Irreversible — real emails
add_newsletter_subscriber	POST	/api/newsletters/:newsletterId/subscribers
remove_newsletter_subscriber	DELETE	/api/newsletters/:newsletterId/subscribers/:subscriberId	Hard-delete
list_newsletter_subscribers	GET	/api/newsletters/:newsletterId/subscribers	Filter by status
get_subscriber_profile	GET	/api/newsletter-subscribers/:subscriberId/profile	Questions + answers + metadata
submit_subscriber_answers	POST	/api/newsletter-subscribers/:subscriberId/answers	Idempotent
get_subscriber_journal	GET	/api/newsletter-subscribers/:subscriberId/journal	Issue history
create_survey	POST	/api/surveys	Body includes datasphereUri
generate_survey_questions	POST	/api/surveys/:pageId/generate-questions	AI writes questions
add_survey_question	POST	/api/surveys/:pageId/questions
update_survey_question	PATCH	/api/surveys/:pageId/questions/:questionId
delete_survey_question	DELETE	/api/surveys/:pageId/questions/:questionId
reorder_survey_questions	POST	/api/surveys/:pageId/questions/reorder
list_survey_questions	GET	/api/surveys/:pageId/questions
get_survey_responses	GET	/api/surveys/:pageId/responses
get_survey_analytics	GET	/api/surveys/:pageId/analytics
export_survey_responses	GET	/api/surveys/:pageId/export/:format	csv or json
link_survey_to_newsletter	POST	/api/newsletters/:newsletterId/generate-survey-codes	Key wiring step
suggest_survey_mappings	POST	/api/newsletters/:newsletterId/suggest-survey-mappings	AI designs metadata fields
create_survey_mapping	POST	/api/newsletters/:newsletterId/survey-mappings	Apply one mapping
list_survey_mappings	GET	/api/newsletters/:newsletterId/survey-mappings	Verify wiring

Mount note: create_newsletter / update_newsletter are at /api/ (not /api/v1/). list_newsletters uses the datasphere URI; all other newsletter tools use the newsletter's DB ID.

---

URL Reference

URL	Purpose
/newsletters/{slug}	Public reader + subscribe page — share this everywhere
/newsletters/{slug}/issues/{issueSlug}	Individual public issue
/newsletters/{slug}/thread	Community discussion thread
/newsletters/{slug}/issues/{issueSlug}/personalized	Authenticated subscriber's personalized view
/s/{uri}/{slug}?code={inviteCode}	Subscriber's personalized survey URL (auto-included in every letter)
/app/{uri}/newsletters/{id}/subscribers	Admin subscriber management — bulk actions, CSV, survey status
/app/{uri}/newsletters/{id}/issues/{issueId}/edit	Issue editor
/app/{uri}/surveys	Survey list for a datasphere
/survey/{uri}/{slug}	Take a survey (public)
/survey/{uri}/{slug}/live	Live results view

---

Error Patterns

Error	Cause	Fix
400 on create_newsletter	Missing name or datasphereId	Both required
400 on draft_newsletter_issue	systemInstructions is empty	Call update_newsletter to set it first
403 on any newsletter write	Not ADMIN or OWNER of the datasphere	Newsletters require admin/owner role
404 on subscriber tools	Wrong subscriberId or newsletter ID	Get IDs fresh from list_newsletter_subscribers
Preview returns generic letter	Personalization not wired	Check: personalizationMode=PERSONALIZED, survey linked, mappings exist, subscriber has answers
PENDING subscribers not receiving	Invited with invite=true, haven't confirmed	Resend invite or add them as ACTIVE directly
private_thread_enabled replies not routing	billing_mode mismatch or capacity exhausted	Check datasphere billing pool; switch to SUBSCRIBER mode if needed
CHECKBOX question only allows one select	allowMultiple is false	Re-create question with allowMultiple: true

---

Activity Digest (Separate System)

The Activity Digest is a per-user, platform-managed email — NOT a newsletter. It auto-scores activity from the user's dataspheres and sends a personalized summary.

Feature	Newsletter	Activity Digest
Scope	Datasphere publication	Per-user, cross-datasphere
Content	Admin/AI-generated	Auto-scored from member activity
Schedule	Datasphere-controlled	User-controlled (daily/weekly/biweekly)
Recipients	DS subscribers	Individual member only

Tools: digest_preview, digest_history, digest_settings. API: GET /api/users/me/activity-digest/preview?dataOnly=true