/work
Execute a plan end-to-end with checks
$ npx -y skills add gmickel/flow-next --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/work
Context preview
What this command does when you run it.
Execute a plan end-to-end with checks
Command definition
work.mdname: work description: Execute a plan end-to-end with checks argument-hint: "<fn-N or idea> [--branch=current|new|worktree] [--review=rp|export|none] [delegate:codex|delegate:local]"
IMPORTANT: This command MUST invoke the skill `flow-next-work`
The ONLY purpose of this command is to call the `flow-next-work` skill. You MUST use that skill now.
**User input:** $ARGUMENTS
Pass the user input to the skill. The skill handles all execution logic.
Repeatable agentic engineering. The workflow layer that turns AI coding agents into a disciplined factory: durable specs, fresh-context workers, adversarial cross-model reviews, receipts. Everything in your repo, zero dependencies. Claude Code · Codex · Cursor · Droid.
Other commands on flow-next.
- /audit
Audit `.flow/memory/` entries against the current codebase — Keep / Update / Consolidate / Replace / Delete / Harden per entry
Open command - /capture
Synthesize the current conversation into a flow-next spec at `.flow/specs/<spec-id>.md` — source-tagged, mandatory read-back, no silent overwrite
Open command - /chart
Decision-map discovery for one oversized/unclear idea before capture - ground, chart a frontier, resolve one decision per invocation, brief for capture
Open command - /guide
Prompt-first router that recommends the smallest sufficient flow-next workflow for the current starting state
Open command - /impl-review
John Carmack-level implementation review via RepoPrompt or Codex
Open command - /interview
Interview & refine a spec, task, or spec file in-depth
Open command