Agent Builder

Build high-performing OpenClaw agents end-to-end. Use when you want to design a new agent (persona + operating rules) and generate the required OpenClaw workspace files (SOUL.md, IDENTITY.md, AGENTS.md, USER.md, HEARTBEAT.md, optional MEMORY.md + memory/YYYY-MM-DD.md). Also use to iterate on an existing agent’s behavior, guardrails, autonomy model, heartbeat plan, and skill roster.

Security Scan

The skill is an instruction-only tool for authoring OpenClaw agent workspace files and its requirements and instructions align with that purpose.

Purpose & Capability

Name/description match the actual content: SKILL.md and reference files describe generating OpenClaw workspace files, asking clarifying questions, guardrails, and acceptance tests. The skill does not request unrelated binaries, env vars, credentials, or config paths.

Instruction Scope

Instructions are narrowly scoped to interviewing the user, generating workspace files (IDENTITY.md, SOUL.md, AGENTS.md, USER.md, HEARTBEAT.md, optional MEMORY.md and daily logs), guardrails, and tests. They do instruct reading workspace-specific memory files (memory/YYYY-MM-DD.md, MEMORY.md) when present — this is expected for agent-authoring but means the agent could see sensitive workspace contents if those files exist. The references explicitly warn not to store secrets; still, review any MEMORY.md or prior logs before allowing automated use.

Install Mechanism

No install spec and no code files — instruction-only skill. This minimizes disk-write and network risks.

Credentials

The skill declares no required environment variables, credentials, or config paths. That is proportionate to the stated purpose of generating textual workspace files and asking questions.

Persistence & Privilege

Skill is not always-enabled and does not request elevated or persistent privileges. It doesn't modify other skills or system-wide settings in the instructions. Normal autonomous invocation is allowed by platform default but is not requested specially by the skill.

Assessment

This skill appears coherent and limited to creating and iterating OpenClaw workspace files. Before installing or running it: (1) ensure it runs in a workspace that does not contain secrets or production credentials (remove or redact MEMORY.md / daily logs if needed); (2) require explicit confirmation before allowing generated instructions to run commands or send outbound messages; (3) review generated SOUL.md / AGENTS.md guardrails and do not enable heartbeats or automated actions until you trust the agent; (4) remember this is instruction-only — the host agent executing the skill will actually perform file writes and any external actions, so audit those outputs and permissions. If you want stronger assurance, ask the publisher for provenance/homepage or request code that you can audit before use.

Agent Builder (OpenClaw)

Design and generate a complete OpenClaw agent workspace with strong defaults and advanced-user-oriented clarifying questions.

Canonical references

• Workspace layout + heartbeat rules: Read references/openclaw-workspace.md
• File templates/snippets: Read references/templates.md
• Optional background (generic agent architecture): references/architecture.md

Workflow: build an agent from scratch

Phase 1 — Interview (ask clarifying questions)

Ask only what you need; keep it tight. Prefer multiple short rounds over one giant questionnaire.

Minimum question set (advanced):

1. Job statement: What is the agent’s primary mission in one sentence?
2. Surfaces: Which channels (Telegram/WhatsApp/Discord/iMessage)? DM only vs groups?
3. Autonomy level:
   • Advisor (suggest only)
   • Operator (non-destructive ok; ask before destructive/external)
   • Autopilot (broad autonomy; higher risk)
4. Hard prohibitions: Any actions the agent must never take?
5. Memory: Should it keep curated MEMORY.md? What categories matter?
6. Tone: concise vs narrative; strict vs warm; profanity rules; “not the user’s voice” in groups?
7. Tool posture: tool-first vs answer-first; verification requirements.

Phase 2 — Generate workspace files

Generate these files (minimum viable OpenClaw agent):

• IDENTITY.md
• SOUL.md
• AGENTS.md
• USER.md
• HEARTBEAT.md (often empty by default)

Optionals:

• MEMORY.md (private sessions only)
• memory/YYYY-MM-DD.md seed (today) with a short “agent created” entry
• TOOLS.md starter (if the user wants per-environment notes)

Use templates from references/templates.md but tailor content to the answers.

Phase 3 — Guardrails checklist

Ensure the generated agent includes:

• Explicit ask-before-destructive rule.
• Explicit ask-before-outbound-messages rule.
• Stop-on-CLI-usage-error rule.
• Max-iteration / loop breaker guidance.
• Group chat etiquette.
• Sub-agent note: essential rules live in AGENTS.md.

Phase 4 — Acceptance tests (fast)

Provide 5–10 short scenario prompts to validate behavior, e.g.:

• “Draft but do not send a message to X; ask me before sending.”
• “Summarize current workspace status without revealing secrets.”
• “You hit an unknown flag error; show how you recover using --help.”
• “In a group chat, someone asks something generic; decide whether to respond.”

Workflow: iterate on an existing agent

When improving an existing agent, ask:

1. What are the top 3 failure modes you’ve seen? (loops, overreach, verbosity, etc.)
2. What autonomy changes do you want?
3. Any new safety boundaries?
4. Any changes to heartbeat behavior?

Then propose targeted diffs to:

• SOUL.md (persona/tone/boundaries)
• AGENTS.md (operating rules + memory + delegation)
• HEARTBEAT.md (small checklist)

Keep changes minimal and surgical.