Skip to content
Back to skills

Standup

ASecurity

Digest of what the user's Superset agents did while they were away, sweeping workspaces, tasks, and agent terminals to report what finished, what needs review, and what's blocked. Use when the user asks "what did my agents do", "what happened while I was away", wants a standup or summary of agent work, or returns after a break.

  • 14,923 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 2, 2026
ai-agentsbash

Works with

  • claude code
  • terminal
  • cli

Security analysis

A100/100

Pro scans all 3 files and shows the line behind each finding

Scanned September 23, 2026

npx -y skills add superset-sh/superset --skill standup --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Standup?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Standup
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/superset-sh-standup/badge)](https://www.skillsdirectory.com/skills/superset-sh-standup)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: standup
description: Digest of what the user's Superset agents did while they were away, sweeping workspaces, tasks, and agent terminals to report what finished, what needs review, and what's blocked. Use when the user asks "what did my agents do", "what happened while I was away", wants a standup or summary of agent work, or returns after a break.
argument-hint: optional timeframe or project
allowed-tools: Bash(superset:*) Bash(bash ${CLAUDE_SKILL_DIR}/scripts/sweep.sh *)
---

# Superset Standup

Answer "what happened while I was away?" from real state, not guesses. Entirely read-only.

## 1. Sweep

Run `bash scripts/sweep.sh` from this skill's directory (`bash ${CLAUDE_SKILL_DIR}/scripts/sweep.sh` in Claude Code). It lists workspaces and tasks, then for each workspace lists its terminals and reads the last screen of every agent terminal, tolerating individual failures. Pass `--host <id>` to sweep a remote host, `--max-lines <n>` to change the per-terminal tail (default 60).

If the script can't run (no bash, no `jq`), do the same by hand: `superset workspaces list --local --json`, `superset tasks list --json`, then per workspace `superset terminals list --workspace <id> --json` and `superset terminals read --workspace <id> --terminal <terminalId> --max-lines 60`.

The last screen of a terminal shows whether the agent finished, asked a question, or errored.

## 2. Classify each workspace

- **Needs you**: agent finished and awaits review, asked a question, or hit a permission prompt or failure
- **In flight**: actively working
- **Blocked**: waiting on something external
- **Stale**: idle with no pending work, a candidate for cleanup

## 3. Report

Lead with what needs the user, one line per item: workspace, agent, state, and the next action. Then in-flight, then completed, then stale-workspace cleanup suggestions. Keep the whole digest scannable: no terminal dumps, quote at most the single relevant line an agent printed.

Offer the digest as a page when someone other than the user will read it, or when they want it
to persist past the scrollback. Write the report you just composed to an `.html` file inside a
workspace, then publish that file: the CLI publishes a document, not your terminal output.

```bash
superset pages publish standup.html --workspace <id> --title "Standup" --label "what shipped overnight"
```

Publishing the same file from the same workspace each morning versions one page instead of
scattering a new one daily, so the history becomes the record of the week. A digest often runs
from outside any workspace, and a publish with no workspace is refused, so pass `--workspace`
explicitly. Offer it; never publish one unasked.

## Rules

Never send input to a terminal, modify tasks, or clean anything up as part of the digest. Offer those as follow-ups and act only when asked.

Files in this skill

  • SKILL.md2.1 KB
  • agents/openai.yaml231 B
  • scripts/sweep.sh2.3 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.