agentleFS
fworks-techfworks-tech/agenthood/skills/mailman/SKILL.md

mailman

Skill2 starsChanged todayMITView on GitHub

Manages message delivery, content scheduling, notification dispatch, and channel management. Use before publishing any scheduled content, when configuring notification pipelines, or when setting up delivery workflows.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 11 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before publishing any scheduled content — to verify delivery pipeline integrity

    line 16
  2. Distinctive:

    When configuring notification systems (email, push, webhook, Slack)

    line 17
  3. Distinctive:

    When a scheduled task failed to execute or a notification wasn't delivered

    line 18
  4. Distinctive:

    When auditing delivery logs for reliability metrics

    line 19
  5. Distinctive:

    Check the delivery manifest: what needs to go where, and by when

    line 25
  6. Distinctive:

    Verify each channel's health:

    line 26
  7. Distinctive:

    If dry-run passes, release the batch with tracking headers

    line 32
  8. Distinctive:

    After delivery, confirm receipt signals — log any failures for retry

    line 33
  9. Distinctive:

    Check the schedule against channel constraints:

    line 38
  10. Distinctive:

    Every delivery attempt records:

    line 61
  11. Distinctive:

    Before a scheduled publish:

    line 102

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

The Mailman

Overview

The Mailman does not create content. It delivers it. Every notification reaches its destination. Every scheduled post publishes on time. Every message appears in every channel it belongs in. The Mailman is the Society's outgoing communications infrastructure — the courier that ensures nothing gets lost in transit, no deadline slips, and no channel goes silent.

When to Use

  • Before publishing any scheduled content — to verify delivery pipeline integrity
  • When configuring notification systems (email, push, webhook, Slack)
  • When a scheduled task failed to execute or a notification wasn't delivered
  • When auditing delivery logs for reliability metrics

Process

Delivery Pipeline Verification

  1. Check the delivery manifest: what needs to go where, and by when
  2. Verify each channel's health:
    • Email: SMTP reachable, queue depth normal, bounce rate below threshold
    • Push: Web Push API endpoint reachable, subscription count matches expected
    • Webhook: Target endpoints respond 200, timeout configs aren't too tight
    • Internal: API keys are valid, rate limits aren't exhausted
  3. Dry-run the batch: simulate delivery without sending live
  4. If dry-run passes, release the batch with tracking headers
  5. After delivery, confirm receipt signals — log any failures for retry

Content Scheduling

  1. Accept the content payload: article body, metadata, target channels, publish time
  2. Check the schedule against channel constraints:
    • Rate limits (API calls per hour, posts per day)
    • Time-of-day preferences (don't post at 3 AM local if it's a personal account)
    • Content size limits per channel
  3. Register scheduled delivery in two places:
    • Local job queue: for immediate execution responsibility
    • Persistent store: for crash recovery (if the scheduler restarts, what still needs to go out?)
  4. At publish time, execute the delivery and log status

Notification Dispatch

  1. Determine the notification type: push, email, in-app, webhook
  2. Route through the appropriate provider:
    • Push: Web Push API (VAPID keys, subscription management)
    • Email: SMTP / SendGrid / SES via transport layer
    • Webhook: HTTP POST with signature verification
    • In-app: Server-Sent Events or WebSocket broadcast
  3. Apply per-channel formatting (HTML for email, markdown for webhook, notification payload for push)
  4. Send with idempotency key — if the same notification is submitted twice, it should only be delivered once
  5. On failure: retry with exponential backoff (1s → 4s → 16s → max 3 retries), then escalate

Delivery Logging & Auditing

Every delivery attempt records:

{
  "id": "dlv_abc123",
  "type": "notification",
  "source": "system-alert",
  "channels": ["email", "webhook"],
  "status": "delivered",
  "results": {
    "email": { "status": "delivered", "latency": 1200 },
    "webhook": { "status": "delivered", "latency": 300 }
  },
  "timestamp": "2026-07-06T14:00:00Z"
}

The Mailman maintains a rolling 7-day delivery log and can answer:

  • What was delivered in the last 24 hours?
  • Which channel has the highest failure rate?
  • Are any scheduled tasks overdue?

Red Flags

  • A scheduled post that did not publish at its target time
  • A notification channel with delivery latency > 30 seconds
  • Delivery logs showing the same task submitted more than 3 times
  • An API key expiring within the next 7 days
  • A webhook endpoint returning non-200 for 3 consecutive attempts

Rationalizations

What you thinkWhat The Mailman knows
"I'll just post it manually"Manual posting forgets channels. Automation remembers all of them.
"The notification went through, I saw it"One success doesn't mean the pipeline is healthy. Check the logs.
"Scheduling a week ahead is risky"Scheduling with a dry-run is safer than last-minute publishing.
"Rate limits won't matter for one post"They matter when you're resubmitting the failed post plus the new one.

Verification

Before a scheduled publish:

  • Delivery manifest is complete — every channel listed
  • All target API keys are valid and not expiring within 7 days
  • Rate limits are respected — no channel exceeds 80% of its hourly quota
  • Dry-run passed — no formatting errors, no missing fields
  • Idempotency keys are set — duplicate submissions won't double-deliver
  • Retry policy is configured — exponential backoff with max 3 attempts
  • Fallback channel exists for critical notifications (email is always the fallback)
  • Delivery log is being written to the configured output

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.