ci-triage
Diagnose a failing GitHub Actions run — find the first real error in the logs, tell a flake apart from a genuine failure, and identify the commit that broke…
Draft a Linear project status update from what actually moved — progress, risks, and the one decision that needs making. Use when someone asks for a project update, a status post, where a project stands, whether it will land on time, or what to tell stakeholders.
$ npx -y skills add superset-sh/superset --skill project-status --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/project-statusContext preview
The summary Claude sees to decide when to auto-load this skill.
Draft a Linear project status update from what actually moved — progress, risks, and the one decision that needs making. Use when someone asks for a project update, a status post, where a project stands, whether it will land on time, or what to tell stakeholders.
name: project-status description: Draft a Linear project status update from what actually moved — progress, risks, and the one decision that needs making. Use when someone asks for a project update, a status post, where a project stands, whether it will land on time, or what to tell stakeholders. argument-hint: project name or URL allowed-tools: mcp__linear__*
A status update is read by people who will not open a single issue. It has to survive being the only thing they read: **is this landing, what changed, and what do you need from them.**
Percent-complete is the number people reach for and the one that lies. A project can sit at 70% for a month while the last three issues get harder. What tells the story is what moved since the previous update: issues completed, issues added, and target dates that shifted.
Fetch the project's issues plus its previous status update. Without the previous one you are writing a summary, not an update, and the reader cannot tell what is new.
Pick one and commit to it. Hedging every sentence is how a project stays green until the week it ships late.
| Call | What justifies it | | --- | --- | | **On track** | Remaining scope fits the remaining time at the current rate | | **At risk** | A dependency, an unknown, or a rate that has to improve to land | | **Off track** | The date will move, or scope has to be cut — say which |
If you cannot tell, the answer is at risk, and the reason is that nobody can tell.
"Backend integration is risky" gives the reader nothing. "The export endpoint is not built, we cannot finish import until it is, and the owner is on the payments migration until the 14th" tells them what to move. A risk is a specific event with a specific consequence and usually a specific person who can prevent it.
Include scope that was added mid-project. Silently absorbed scope is why the date moves, and it is invisible unless the update names it.
Most updates end with nothing to do, so nothing is done. If a decision, a person, or a descope would help, state it as the last line with a name and a date attached. One ask is better than three, because three get triaged and one gets answered.
Print the draft in full and stop. This skill drafts; publishing is the user's call. A status update is read by stakeholders and enters the project's permanent history, and a wrong trajectory or a risk named for the wrong person is not something a follow-up correction undoes — so an unreviewed post is worse than a late one.
Once they approve it, write it on the project so it lives next to the work rather than in a chat message that is gone by the next morning. Keep it to a few short paragraphs: trajectory, what moved, risks, the ask. If they asked only for a draft, stop after printing it.
the one update that matters gets skipped too.
Superset is an agentic IDE to orchestrate 100+ coding agents in parallel. Run any agent with your own subscription.
Diagnose a failing GitHub Actions run — find the first real error in the logs, tell a flake apart from a genuine failure, and identify the commit that broke…
Triage a GitHub issue into something actionable — reproduce the claim, find duplicates, judge severity, and apply labels, milestone, and assignee. Use when the…
Find and merge duplicate Linear issues — group reports of the same underlying bug, pick the survivor, and move the evidence across. Use when the backlog has…
Turn a rough report into a Linear issue someone can pick up — reproduce the claim, check for duplicates, and fill in team, priority, and labels. Use when the…
Personalized audit that teaches the advanced Superset features the user isn't using yet (automations, parallel agents, tasks, multi-host, terminal remote…
Turn a recurring chore into a Superset automation. Drafts the agent prompt, confirms schedule and target, creates it with the CLI, and reviews the first run…