hub-status — Claude Code skill: AgentHub session status
Shows DAG state, agent progress and branch status for an AgentHub session. Handy when several agents share one task and you need to see who is where.
/plugin marketplace add alirezarezvani/claude-skills
/plugin install agenthub@claude-code-skillsWhat hub-status does
The skill runs as the slash command /hub:hub-status. With no argument it reports the latest session; pass an id like 20260317-143022 to target a specific one.
Three scripts do the work: session_manager.py prints the session overview, dag_analyzer.py walks the task graph, and board_manager.py reads recent board entries. You get a table of agents with their branches (hub/<session>/agent-N/attempt-1), commit counts, frontier status and last update time, followed by board activity tagged [progress] or [results].
It sits on top of AgentHub, so there is nothing to show until a session is running. When every agent has posted results, the skill points you at /hub:eval to rank them; when some are still working, it lists which are done and which are not.
What it helps with
- see how many commits each agent made
- tell finished agents from ones still running
- read the latest board progress entries
- check a specific session by its id
- decide when to rank the results
Example prompts for the agent
- /hub:hub-status
- /hub:hub-status 20260317-143022
- how are the AgentHub agents doing, show me the status
How to install hub-status
1Plugin marketplace
/plugin marketplace add alirezarezvani/claude-skills/plugin install agenthub@claude-code-skillsThis installs the whole agenthub plugin with all its skills
2Skills CLI (Vercel)
npx skills add alirezarezvani/claude-skills --skill hub-status -a claude-code3Manually: put the files in place
~/.claude/skills/hub-status/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/hub-status/
Manually: put the files in place
.agents/skills/hub-status/SKILL.mdFor all projects use ~/.agents/skills/
Manually: put the files in place
.cursor/skills/hub-status/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install alirezarezvani/claude-skills hub-status2Manually: put the files in place
.github/skills/hub-status/SKILL.mdPlatform CLI
gemini skills install https://github.com/alirezarezvani/claude-skills --path engineering/agenthub/skills/hub-statusManually: put the files in place
.devin/skills/hub-status/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about hub-status
How do I install hub-status in Claude?
Through the marketplace: run /plugin marketplace add alirezarezvani/claude-skills, then /plugin install agenthub@claude-code-skills.
What does hub-status do?
It prints the state of an AgentHub session: agent branches, commit counts, frontier status and recent board updates.
What does the skill need to work?
A running AgentHub session. Without one there is nothing to read, since the skill inspects that session's branches and board.
Source text
SKILL.md · MIT
---
name: "hub-status"
description: "Show DAG state, agent progress, and branch status for an AgentHub session. Use when the user runs /hub:hub-status or asks how the AgentHub agents are doing."
metadata:
command: /hub:hub-status
---
# /hub:hub-status — Session Status
Display the current state of an AgentHub session: agent branches, commit counts, frontier status, and board updates.
## Usage
```
/hub:hub-status # Status for latest session
/hub:hub-status 20260317-143022 # Status for specific session
```
## What It Does
1. Run session overview:
```bash
python {skill_path}/scripts/session_manager.py --status {session-id}
```
2. Run DAG analysis:
```bash
python {skill_path}/scripts/dag_analyzer.py --status --session {session-id}
```
3. Read recent board updates:
```bash
python {skill_path}/scripts/board_manager.py --read progress
```
## Output Format
```
Session: 20260317-143022 (running)
Task: Optimize API response time below 100ms
Agents: 3 | Base: dev
AGENT BRANCH COMMITS STATUS LAST UPDATE
agent-1 hub/20260317-143022/agent-1/attempt-1 3 frontier 2026-03-17 14:35:10
agent-2 hub/20260317-143022/agent-2/attempt-1 5 frontier 2026-03-17 14:36:45
agent-3 hub/20260317-143022/agent-3/attempt-1 2 frontier 2026-03-17 14:34:22
Recent Board Activity:
[progress] agent-1: Implemented caching, running tests
[progress] agent-2: Hash map approach working, benchmarking
[results] agent-2: Final result posted
```
Example output for a content task:
```
Session: 20260317-151200 (running)
Task: Draft 3 competing taglines for product launch
Agents: 3 | Base: dev
AGENT BRANCH COMMITS STATUS LAST UPDATE
agent-1 hub/20260317-151200/agent-1/attempt-1 2 frontier 2026-03-17 15:18:30
agent-2 hub/20260317-151200/agent-2/attempt-1 2 frontier 2026-03-17 15:19:12
agent-3 hub/20260317-151200/agent-3/attempt-1 1 frontier 2026-03-17 15:17:55
Recent Board Activity:
[progress] agent-1: Storytelling angle draft complete, refining CTA
[progress] agent-2: Benefit-led draft done, testing urgency variant
[results] agent-3: Final result posted
```
## After Status
If all agents have posted results:
- Suggest `/hub:eval` to rank results
If some agents are still running:
- Show which are done vs in-progress
- Suggest waiting or checking again later
© Alireza Rezvani · MIT · alirezarezvani/claude-skills
