Monitor Digest Notifier — one readable daily message from all your Apify monitors

You schedule five monitors (SAM.gov opportunities, SEC filings, federal court cases, building permits, trademark filings…) and each one fires its own raw JSON webhook. Nobody reads five JSON blobs.

This actor collects the output of every monitor run from the last 24 hours, renders one digest grouped by monitor, and delivers it to Slack, Discord, Microsoft Teams and/or email. It also stores the digest as Markdown in the key-value store and as a structured record in the dataset, so it works as a plain "what changed today" report even with no channel configured.

Price: $0.05 per digest (plus $0.005 per start). One run = one digest = one charge, however many monitors it summarizes.

Auto-discovery of your recent monitor runs needs a personal API token in apifyToken (actor runs use a limited-permission token that cannot list runs). With explicit datasetIds or runIds no token is required.

What the digest looks like

# Daily monitor digest
2026-09-28 13:45 UTC · 2 monitors · 7 new records · last 24h

## SAM.gov Opportunity Monitor
Run 2026-09-28 13:30 UTC · 3 new, 1 updated · [open run](https://console.apify.com/actors/runs/…)

- [Cybersecurity Support Services for Regional Offices](https://sam.gov/opp/a1/view) — 2026-09-27 · DEPT OF DEFENSE
- [Zero Trust Architecture Assessment & Roadmap](https://sam.gov/opp/a2/view) — 2026-09-27 · HOMELAND SECURITY, DEPARTMENT OF · updated
- [SOC-as-a-Service for VA Medical Centers](https://sam.gov/opp/a3/view) — 2026-09-28 · VETERANS AFFAIRS, DEPARTMENT OF
- _[+1 more](https://console.apify.com/storage/datasets/…)_

## Federal Court Case Monitor
Run 2026-09-28 13:40 UTC · 1 new case, 1 new filing, 1 new opinion · [open run](…)

- [Smith v. Amazon.com, Inc.](https://www.courtlistener.com/docket/70001/…) — 2026-09-26 · District Court, W.D. Washington · new case
- [In re Tesla, Inc. Securities Litigation](https://www.courtlistener.com/docket/69120/…) — 2026-08-02 · District Court, N.D. California · new filing
- [Tesla, Inc. v. Louisiana Automobile Dealers Association](https://www.courtlistener.com/opinion/88801/…) — 2026-09-25 · Court of Appeals for the Fifth Circuit · new opinion

Slack gets the same content as Block Kit sections with clickable links, Discord as messages of at most 2,000 characters (split automatically, link previews suppressed), Teams as an Adaptive Card, email as HTML with a plain-text alternative.

How it finds your monitors

Three ways, in order of precedence:

1. datasetIds — explicit dataset IDs or username~datasetName.
2. runIds — explicit run IDs; their default datasets are used and the actor name, run time and SUMMARY record are read from the run.
3. Auto-discovery (default, empty input) — lists your own successful runs from the last lookbackHours (24 by default) and keeps every run whose actor name contains "monitor". Works with any actor, yours or from the Store, as long as the name contains "monitor".

With groupBy: "monitor" (default) several runs of the same actor in the window are merged into one section; groupBy: "run" keeps one section per run.

Every record is rendered by a generic field picker, so any monitor works without configuration:

Column	Taken from the first non-empty of
title	title, headline, caseName, markText, businessName, establishmentName, companyName, advertiserName, recipientName, name, address, else the first text field
link	url, samUrl, detailUrl, courtListenerUrl, tsdrUrl, oshaUrl, grantsUrl, bodaccUrl, usaspendingUrl, filingUrl, profileUrl, bidUrl, adUrl, permitUrl, listingUrl, else any *Url field
date	first field ending in Date, then date*, then *At (run-time stamps like firstSeenAt skipped)
context	agency, courtName, court, formType, noticeType, productName, filingType, competitor, company, insurer, state, city, category
change	changeType (counted per monitor: "3 new, 1 updated", "1 new case, 1 new filing"…)

Channel setup

Set channels to a JSON object with any of these keys:

{
  "slackWebhookUrl": "https://hooks.slack.com/services/T000/B000/XXXX",
  "discordWebhookUrl": "https://discord.com/api/webhooks/123/abc",
  "teamsWebhookUrl": "https://prod-00.westus.logic.azure.com:443/workflows/…",
  "emailTo": "ops@example.com, legal@example.com"
}

• Slack — create an Incoming Webhook app (api.slack.com/apps → Incoming Webhooks → Add to workspace) and paste the URL.
• Discord — channel settings → Integrations → Webhooks → New webhook → Copy URL.
• Microsoft Teams — channel → Workflows → "Post to a channel when a webhook request is received" (or a legacy Incoming Webhook connector). Both accept the Adaptive Card this actor sends.
• Email — sent through Apify's own apify/send-mail actor from your account; no SMTP setup. Comma-separate multiple addresses.

Set dryRun: true to render everything and log the exact payload for each channel without sending. Use includeLinks: false for channels where links are unwanted.

Chaining it after your monitors

Option A — schedule (simplest). Put your monitors on a schedule (e.g. 0 6 * * *) and schedule this actor 30 minutes later (30 6 * * *) with the default input. It picks up everything that finished in the last 24 hours.

Option B — webhook per monitor (instant). On each monitor actor, add a webhook with event ACTOR.RUN.SUCCEEDED and request URL

https://api.apify.com/v2/acts/<your-username>~monitor-digest-notifier/runs?token=<your-token>

and payload template

{ "runIds": ["{{resource.id}}"], "title": "{{resource.actId}} finished", "channels": { "slackWebhookUrl": "https://hooks.slack.com/…" } }

You then get one message per monitor run, formatted, instead of raw JSON.

Option C — call it from your own actor. await Actor.call('<your-username>/monitor-digest-notifier', { datasetIds: ['username~my-dataset'], channels: { … } }).

Input

Field	Default	Meaning
datasetIds	[]	Dataset IDs or username~name to include
runIds	[]	Run IDs whose default datasets to include
lookbackHours	24	Auto-discovery window when neither list is given
channels	{}	slackWebhookUrl, discordWebhookUrl, teamsWebhookUrl, emailTo
title	Daily monitor digest	Heading and email subject
maxItemsPerMonitor	10	Rows listed per monitor (the full count is always shown)
groupBy	monitor	monitor merges runs of the same actor; run keeps them apart
includeLinks	true	Link records and runs
dryRun	false	Render and log payloads without sending
maxItems	10	Max monitor runs included, most recent first

Output

One dataset record per run:

{
  "title": "Daily monitor digest",
  "generatedAt": "2026-09-28T13:45:00.000Z",
  "lookbackHours": 24,
  "monitorCount": 2,
  "totalRecords": 7,
  "monitors": [
    {
      "name": "SAM.gov Opportunity Monitor",
      "slug": "sam-gov-opportunity-monitor",
      "runId": "abc123",
      "runIds": ["abc123"],
      "datasetId": "def456",
      "runAt": "2026-09-28T13:30:00.000Z",
      "newCount": 4,
      "countsByChangeType": { "new": 3, "updated": 1 },
      "summary": { "monitorId": "default", "newCount": 3, "updatedCount": 1, "scanned": 212 },
      "sample": [
        { "title": "Cybersecurity Support Services for Regional Offices", "url": "https://sam.gov/opp/a1/view", "date": "2026-09-27", "meta": "DEPT OF DEFENSE", "changeType": "new" }
      ]
    }
  ],
  "markdown": "# Daily monitor digest…",
  "text": "DAILY MONITOR DIGEST…",
  "delivery": { "slack": { "ok": true, "sent": 1 }, "email": { "ok": true, "runId": "…", "status": "SUCCEEDED" } }
}

The Markdown digest is also saved to the run's key-value store as DIGEST.md. If no monitor runs are found, the actor still emits one record saying so, so scheduled runs never come back empty.

Notes

• Auto-discovery uses the run's own token, so it only sees runs in the account that runs this actor.
• Up to 1,000 records per dataset are read for counting; maxItemsPerMonitor controls how many are listed.
• Slack sections are kept under 3,000 characters and messages under 45 blocks; Teams cards under ~20 KB; anything larger is split into several messages.