Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Route agent notifications to specific channels by type — prevent alert fatigue from single-channel flooding
.claude/skills/github-notification-routing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 124% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 102% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -15% | 0% |
When a Squad grows beyond a few agents, notifications flood a single channel — failure alerts drown in daily briefings, tech news buries security findings, and everything gets ignored. This is the pub-sub problem: a single message queue for everything is a recipe for missed alerts.
The fix is topic-based routing: agents tag notifications with a channel type, and a routing function sends them to the appropriate destination.
Trigger symptoms:
Define a .squad/teams-channels.json (or equivalent) mapping notification types to channel identifiers:
json{ "teamId": "your-team-id", "channels": { "notifications": "squad-alerts", "tech-news": "tech-news", "security": "security-findings", "releases": "release-announcements", "daily-digest": "daily-digest" } }
Place this in .squad/ (git-tracked, shared across the team). For platforms that use channel IDs instead of names (Teams, Slack), store the resolved ID alongside the name to avoid name-collision bugs:
json{ "channels": { "notifications": { "name": "squad-alerts", "id": "channel-id-opaque-string" } } }
Agents prefix their output with CHANNEL:<type> to signal where the notification should go:
CHANNEL:security
Worf found 3 new CVEs in dependency scan: lodash@4.17.15, minimist@1.2.5bashdispatch_notification() { local raw_output="$1" local channel="notifications" # default if echo "$raw_output" | grep -qE '^CHANNEL:[a-z][a-z0-9-]*'; then channel=$(echo "$raw_output" | head -1 | cut -d: -f2) raw_output=$(echo "$raw_output" | tail -n +2) fi send_notification --channel "$channel" --message "$raw_output" }
The routing layer is provider-agnostic. Plug in your platform adapter:
.squad/notify-adapter.sh # Teams / Slack / Discord / webhook -- swappableThe routing config and CHANNEL: tags never change. Only the adapter changes per deployment.
Never send all notification types to one channel:
send_notification --channel "general" --message "$anything"Never use display names as identifiers (name collision risk):
send_to_team --name "Squad" --channel "notifications"Resolve channel IDs once at setup. Use IDs at runtime.
This is pub-sub with topic routing -- the same principle as Kafka topics, RabbitMQ routing keys, and AWS SNS topic filtering. Route by type. Each consumer subscribes to the topics it cares about.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 37,097 | 16,969 | -54% | 1 | 1 | 0% | 1,779 | 3,984 | +124% | 0 | 0 | — |
case-02 | fail→pass | 14,005 | 10,762 | -23% | 1 | 1 | 0% | 2,592 | 2,551 | -2% | 0 | 0 | — |
case-03 | fail→pass | 43,634 | 16,895 | -61% | 1 | 1 | 0% | 1,933 | 3,898 | +102% | 0 | 0 | — |
case-04 | pass→pass | 9,424 | 7,524 | -20% | 1 | 1 | 0% | 1,811 | 1,922 | +6% | 0 | 0 | — |
case-05 | pass→pass | 11,411 | 9,958 | -13% | 1 | 1 | 0% | 1,759 | 2,483 | +41% | 0 | 0 | — |
case-06 | pass→pass | 9,442 | 6,759 | -28% | 1 | 1 | 0% | 1,483 | 1,611 | +9% | 0 | 0 | — |
case-07 | fail→pass | 8,467 | 5,137 | -39% | 1 | 1 | 0% | 1,246 | 1,283 | +3% | 0 | 0 | — |
case-08 | fail→pass | 13,847 | 7,410 | -46% | 1 | 1 | 0% | 2,446 | 2,075 | -15% | 0 | 0 | — |
case-09 | pass→pass | 47,917 | 15,129 | -68% | 1 | 1 | 0% | 2,919 | 3,292 | +13% | 0 | 0 | — |
case-10 | fail→pass | 17,258 | 15,314 | -11% | 1 | 1 | 0% | 3,124 | 3,466 | +11% | 0 | 0 | — |
case-11 | fail→fail | 13,925 | 8,814 | -37% | 1 | 1 | 0% | 2,352 | 2,099 | -11% | 0 | 0 | — |
case-12 | fail→pass | 12,471 | 4,372 | -65% | 1 | 1 | 0% | 1,971 | 1,409 | -29% | 0 | 0 | — |
case-13 | pass→pass | 7,187 | 7,490 | +4% | 1 | 1 | 0% | 1,176 | 1,770 | +51% | 0 | 0 | — |
case-14 | pass→pass | 6,436 | 3,624 | -44% | 1 | 1 | 0% | 980 | 1,260 | +29% | 0 | 0 | — |
case-15 | pass→pass | 16,240 | 6,775 | -58% | 1 | 1 | 0% | 1,531 | 1,771 | +16% | 0 | 0 | — |
case-16 | fail→fail | 30,213 | 5,085 | -83% | 1 | 1 | 0% | 2,434 | 1,336 | -45% | 0 | 0 | — |
case-17 | fail→fail | 14,248 | 9,461 | -34% | 1 | 1 | 0% | 2,287 | 2,283 | -0% | 0 | 0 | — |
case-18 | fail→pass | 13,603 | 6,189 | -55% | 1 | 1 | 0% | 2,272 | 1,675 | -26% | 0 | 0 | — |
case-19 | fail→pass | 19,778 | 11,195 | -43% | 1 | 1 | 0% | 2,318 | 2,410 | +4% | 0 | 0 | — |
case-20 | pass→pass | 19,490 | 17,465 | -10% | 1 | 1 | 0% | 2,834 | 3,288 | +16% | 0 | 0 | — |
case-21 | fail→pass | 10,915 | 12,424 | +14% | 1 | 1 | 0% | 2,112 | 1,792 | -15% | 0 | 0 | — |
case-22 | pass→pass | 8,217 | 23,197 | +182% | 1 | 1 | 0% | 1,296 | 1,623 | +25% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 20 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +45 percentage points is the difference between those two pass rates over the 20 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.