Claude Code: automating my support rotation

On my team we run a weekly support rotation. Whoever is on duty spends a chunk of every day doing the same three things: triaging new issues in the error tracker, working through support tickets on the task board, and skimming an alerts channel in Slack for anything that's started misbehaving.

None of it is hard. All of it is tedious, context-heavy, and exactly the kind of thing you half-do at 4pm on a Thursday. Three different tools, three different mental models, and the real work — "is this new, is this spiking, does this need a ticket?" — only starts after you've collected everything.

So I handed the collecting to Claude Code.

One command, three sources

The shape that worked is not "one giant prompt." It's one orchestrator skill that knows the daily routine, delegating to small specialist skills that each know how to do exactly one thing.

The orchestrator is dumb on purpose. It says, in effect:

1. Pull unresolved issues assigned to my team from the error tracker (last 14 days), surface anything new or spiking.
2. List support tickets created since the last check, grouped by status.
3. Read the alerts channel since yesterday, one line per alert, each ending in a concrete next action.
4. Combine all three into a single dated digest, newest and loudest first.

Each of those steps points at a specialist — a ticket reader, a categorizer, an order-debugger — that owns the messy details. The orchestrator just sequences them and formats the result. When I say "do the support check", it runs all three top-to-bottom and hands me one consolidated digest instead of three browser tabs.

Why split it into many small skills

The temptation is to cram everything into one mega-skill. Don't. The split pays off because:

• Each skill stays simple enough to trust. A skill that only reads one ticket and renders it as clean markdown is easy to get right and easy to reuse.
• They compose. The same ticket-reader feeds the daily support check, a monthly "group these tickets and find patterns" analysis, and one-off debugging. Write it once, call it from everywhere.
• The orchestrator reads like a checklist. Anyone on the team can open it and see exactly what "support duty" means this week — it's documentation that also happens to execute.

This is the bit worth stealing even if you never touch support tickets: an orchestrator skill that fans out to single-purpose specialist skills. It's the same instinct as small functions and one job per module, just pointed at your daily ops.

What I did not automate

The agent gathers and drafts. It does not get to decide. It proposes a verdict per issue — triage / monitor / noise — and a "today's focus" shortlist, but a human still makes the call on what's a real bug versus what's expected noise. The win isn't that a robot does support; it's that I show up to the judgment part with everything already collected, deduplicated, and linked.

Automate the gathering. Keep the judgment. That line is where this stays useful instead of dangerous.