| name | board |
| description | Workspace dashboard: lists every active framework run across all primitives, highlights blockers (stuck items, exhausted auto-continue caps), shows cumulative cost, and suggests the next action. The natural command to run after re-opening Claude Code in a workspace with in-flight runs, or whenever the user wants a one-shot overview of "what's going on".
USE this skill when:
- the user has just (re)opened the workspace and wants to know what is in flight;
- the user says "what's going on", "show me everything", "any active runs", "where are we", "any pipelines running";
- the user wants a one-shot health check + suggested next action.
DON'T use this to inspect a specific run in depth (use `/agentflow:inspect show <id>` or `tree <id>`), or
to modify state (board is read-only). Explicit: `/agentflow:board` or `/agentflow:board --json`.
|
| allowed-tools | Bash, Read |
| argument-hint | [--json] [--no-failed] |
| disable-model-invocation | false |
/agentflow:board
Surface the workspace's current state in one screen. Read-only, no mutation.
node "${CLAUDE_PLUGIN_ROOT}/dist/inspect.js" board [--json] [--no-failed]
What it shows
- Counts:
N active, M done, K failed
- Cumulative cost: total tokens + USD recorded across all runs
- ACTIVE section: every non-terminal run with its primitive-specific progress (e.g.
foreach: 12/30, pipe: 2/4 stages, iterate: 3/10 iters), plus the parent run if it is a /agentflow:pipe child
- BLOCKERS section: runs that need manual attention — items stuck
in_progress from a dead session, auto_continues cap exhausted, etc.
- FAILED section (top 5; suppress with
--no-failed): runs that ended in failed status, with error preview
- Suggested next actions: concrete commands the user can run —
pipe drive, inspect tree, state/foreach.js reset, etc.
Typical usage
After (re)opening Claude Code in this workspace:
user: /agentflow:board
assistant: (calls the CLI, surfaces the output)
If something is active, sending any next message will trigger the Stop hook at end-of-turn and auto-resume the work. /agentflow:board does not resume anything itself — it only reports.
Important rules
- Read-only: never mutates state.
- To ACT on what board shows, use
/agentflow:runs — the writable sibling: pause/resume one job or the whole engine, set priority (queue order), delete a run, or clean up finished ones. Board reports; runs controls.
- Different from
/agentflow:inspect runs: /agentflow:board is for the "session start" overview with suggestions; /agentflow:inspect runs is the bare tabular listing. /agentflow:board is opinionated, /agentflow:inspect is plumbing.
- Always safe to run: no flags, no risk. Use it freely.