Monitoring
Add and remove sources, create alert rules, and choose how notifications reach you.
Monitoring at /app/monitoring is where you tell HermesAI what to watch and how to surface it.
Three things you do here
Pick monitored sources
Choose from the curated catalog or add custom feeds (RSS, sitemap).
Create alert rules
When a story matches certain criteria, you want to know about it sooner.
Configure delivery channels
In-app, email, or webhook delivery for alerts.
Monitored sources
Browse the catalog
HermesAI ships a curated source catalog — international wires, tier 1 outlets, tier 2 nationals, regional digital. Filter by region, language, category, or tier.
Add what you want HermesAI to watch
Sources you add are added to your tenant's monitored set. The global catalog stays the same; you're just expressing your preferences.
Add custom feeds
Got an RSS, sitemap, or other feed not in the catalog? Add it as a custom source. HermesAI starts watching as soon as the feed validates.
Remove anything noisy
Sources are easy to remove. Coverage from a removed source stops feeding the cluster engine immediately.
Plan-aware caps
Plans have caps on how many sources you can monitor at once. The current cap is shown in the monitoring screen. See Billing for plan-specific limits.
Alert rules
Rules are how you tell HermesAI: "wake me up when a story matches this".
Define what to match
Combine any of: specific sources, categories, keywords, minimum source count, breaking-only mode.
Choose alert mode
- Instant — alert as soon as the cluster matches.
- Digest — bundled into a periodic digest.
- Both — instant for high-confidence matches, digest for the rest.
Choose channels
In-app (always), email, or webhook delivery. You can pick more than one.
Save and test
Save the rule and trigger a test fire to confirm channels work. Webhook test sends a sample payload to your endpoint.
Rule shapes that work
Match by category + keyword. Useful for: "alert me whenever there's coverage of [topic] in [region]".
Example: category = politics, keywords = EU summit, min sources = 2.
Match only stories flagged as breaking by the cluster engine. Filters out background coverage.
Example: breaking-only = true, alert mode = instant.
Only alert when a minimum number of sources cover the same event. Cuts noise from single-source rumours.
Example: min sources = 3, source tier = wire OR tier 1.
Watch a specific list of sources. Useful when you trust a few outlets and want to see anything they publish.
Example: sources = [Reuters, AP, AFP], alert mode = digest.
Alert delivery channels
In-app
Always on. Notifications surface in the workspace bell + a notifications log.
Sent to the address on the user account. Digest mode bundles.
Webhook
HMAC-signed POST to an endpoint you configure. Useful for forwarding to Slack, Teams, or internal tooling.
Alert history
Every fire is logged with what matched, when, and which channels delivered. Visible in /app/monitoring → Alert history.
When something feels off (too noisy, too quiet, missed a story), the history is your first stop.
Plan-aware limits
| Plan | Monitoring rules | Source cap |
|---|---|---|
| Individual Free | 0 | small |
| Individual Plus | 5 | medium |
| Individual Pro | unlimited | larger |
| Business Plus | 5 | medium (shared) |
| Business Pro | unlimited | larger (shared) |
| Enterprise | unlimited | contract |
→ See Billing for current numbers.