Create Agents Wizard
Follow this workflow to create multiple agents with low user effort.
1) Confirm scope
- Ask for total agent count
N.
- Ask each agent
id (lowercase letters/digits/hyphens only).
- Ask each workspace path (default
~/.openclaw/workspace-<id>).
- Ask scaffold mode:
- Standard (default): 6 files
AGENTS.md, SOUL.md, IDENTITY.md, BOOTSTRAP.md, USER.md, STYLE.md
- Fast mode: 2 files only
- If user does not specify, use Standard by default.
2) Process one agent at a time
For each agent, collect info and draft files in fixed order.
- Ask focused questions (use
references/question-bank.md if available).
- Produce an improved draft from user input.
- Ask explicit confirmation:
Confirm write / Revise.
- Write only after confirmation; otherwise keep iterating.
Rules:
- Do not dump a long questionnaire; ask in small rounds (2–4 key questions each).
- Preserve user intent; do not change direction without consent.
- Respect selected mode (Standard 6 files or Fast 2 files).
- If user requests partial scaffolding, allow writing only the confirmed subset and continue/stop as requested.
3) Write files and create agent
After required files for one agent are confirmed:
- If agent does not exist, create it:
openclaw agents add <id> --workspace <path>
- Write approved files into workspace root.
- Optional batch helper:
scripts/scaffold.sh --agent <id> --workspace <path> --from <approved_dir> --force
- Optional identity setup:
openclaw agents set-identity --workspace <path> --from-identity
If user wants docs only, skip CLI config changes.
4) Final summary
When all agents are done, report:
- Created agents (id + workspace)
- Written file list per agent
openclaw agents list output (if CLI available)- Next-step suggestions (skills, routing, default models)
5) Quality bar
AGENTS.md: operation rules, safety boundaries, heartbeat/external-action policySOUL.md: values, working principles, boundaries, styleIDENTITY.md: name, type, vibe, emoji, avatar conventionBOOTSTRAP.md: first-conversation script and initialization stepsUSER.md: addressing, timezone, preferences, taboos, work/life contextSTYLE.md: tone, length preference, banned words, output format preference
6) Failure handling
- Conflicting user inputs: point out conflicts and offer 2 concrete revision options.
- Too-short answers: provide a minimal viable template, then request missing details.
- Scope change mid-way: reconfirm scope, then continue serial processing.
- Write failure: report exact path + error and provide retry command.
1---2name: create-agents-wizard3description: Guided creation of multiple OpenClaw agents and their workspace templates. Use when users ask to create agents in batch, configure new agents, or customize AGENTS/SOUL/IDENTITY/BOOTSTRAP/USER/STYLE files. The workflow confirms scope, collects preferences agent-by-agent, drafts and confirms files, writes to disk, and summarizes results.4---56# Create Agents Wizard78Follow this workflow to create multiple agents with low user effort.910## 1) Confirm scope11121. Ask for total agent count `N`.132. Ask each agent `id` (lowercase letters/digits/hyphens only).143. Ask each workspace path (default `~/.openclaw/workspace-<id>`).154. Ask scaffold mode:16 - **Standard (default):** 6 files17 - `AGENTS.md`, `SOUL.md`, `IDENTITY.md`, `BOOTSTRAP.md`, `USER.md`, `STYLE.md`18 - **Fast mode:** 2 files only19 - `AGENTS.md`, `SOUL.md`205. If user does not specify, use **Standard** by default.2122## 2) Process one agent at a time2324For each agent, collect info and draft files in fixed order.25261. Ask focused questions (use `references/question-bank.md` if available).272. Produce an improved draft from user input.283. Ask explicit confirmation: `Confirm write / Revise`.294. Write only after confirmation; otherwise keep iterating.3031Rules:32- Do not dump a long questionnaire; ask in small rounds (2–4 key questions each).33- Preserve user intent; do not change direction without consent.34- Respect selected mode (Standard 6 files or Fast 2 files).35- If user requests partial scaffolding, allow writing only the confirmed subset and continue/stop as requested.3637## 3) Write files and create agent3839After required files for one agent are confirmed:40411. If agent does not exist, create it:42 - `openclaw agents add <id> --workspace <path>`432. Write approved files into workspace root.44 - Optional batch helper:45 - `scripts/scaffold.sh --agent <id> --workspace <path> --from <approved_dir> --force`463. Optional identity setup:47 - `openclaw agents set-identity --workspace <path> --from-identity`4849If user wants docs only, skip CLI config changes.5051## 4) Final summary5253When all agents are done, report:5455- Created agents (id + workspace)56- Written file list per agent57- `openclaw agents list` output (if CLI available)58- Next-step suggestions (skills, routing, default models)5960## 5) Quality bar6162- `AGENTS.md`: operation rules, safety boundaries, heartbeat/external-action policy63- `SOUL.md`: values, working principles, boundaries, style64- `IDENTITY.md`: name, type, vibe, emoji, avatar convention65- `BOOTSTRAP.md`: first-conversation script and initialization steps66- `USER.md`: addressing, timezone, preferences, taboos, work/life context67- `STYLE.md`: tone, length preference, banned words, output format preference6869## 6) Failure handling7071- Conflicting user inputs: point out conflicts and offer 2 concrete revision options.72- Too-short answers: provide a minimal viable template, then request missing details.73- Scope change mid-way: reconfirm scope, then continue serial processing.74- Write failure: report exact path + error and provide retry command.
dvcrn