Openclaw Default Agent Backstory

Build, bootstrap, and maintain a stable OpenClaw default agent identity by interviewing the user, updating IDENTITY.md, and seeding core context files (AGENT...

Security Scan

The skill is an instruction-only helper that asks identity questions and creates/updates workspace backstory files — its requirements and behavior are consistent with that purpose.

Purpose & Capability

Name/description (agent backstory, bootstrap identity) match the actual instructions: ask questions, synthesize answers, and create/update core context files (AGENTS.md, IDENTITY.md, SOUL.md, MEMORY.md, etc.). No unrelated credentials, binaries, or external services are requested.

Instruction Scope

SKILL.md instructs reading and writing workspace files and asking five focused questions. It does not direct the agent to read unrelated system files, fetch data from external endpoints, or access hidden env vars. It specifies preservation of user-authored content and clear limits on invented personal history.

Install Mechanism

No install spec or code files to write/execute — instruction-only skill (lowest install risk).

Credentials

No environment variables, credentials, or config paths are requested. Files touched are limited to standard workspace context files documented in the skill; these are proportionate to the stated purpose.

Persistence & Privilege

Skill is not always-enabled and does not request elevated platform privileges. Its behavior is limited to creating/updating workspace files (its intended scope).

Assessment

This skill is internally coherent and appears safe to use, but it will read and write files at your workspace root (IDENTITY.md, AGENTS.md, MEMORY.md, etc.). Before running: 1) back up or version-control your workspace so you can review diffs; 2) avoid pasting secrets or sensitive credentials into answers during the interview; 3) confirm any automated file updates before committing changes; and 4) if you want stricter control, ask the agent to show proposed file contents and require your explicit approval before writing.

Use this skill to define or refine a default OpenClaw agent identity and its supporting context files.

Reference model: OpenClaw context and bootstrap behavior from https://docs.openclaw.ai/concepts/context.

Core Rules

• Ask five questions every time unless the user explicitly asks to skip, shorten, or focus.
• Use ranked questions from references/question-bank.md.
• Prefer unanswered or weakly defined areas first.
• Keep content practical and stable for daily use.
• Preserve user-authored content unless the user asks for replacement.
• If user asks for a default memory policy, apply the Primary Daily Driver policy in references/bootstrap-core-files.md.

Mode Selection

Select one mode before writing files:

1. bootstrap mode:

• Use when user asks to bootstrap/onboard/start fresh.
• Use when core files are missing or mostly placeholders.
• Create or refresh baseline content for all core context files.

2. identity-refresh mode:

• Use when user asks for personality/backstory updates only.
• Update IDENTITY.md sections without rewriting unrelated core files.

Core Context Files

In bootstrap mode, ensure these files exist at workspace root with usable starter content:

• AGENTS.md
• SOUL.md
• TOOLS.md
• IDENTITY.md
• USER.md
• HEARTBEAT.md
• BOOTSTRAP.md

Use guidance and starter structure in references/bootstrap-core-files.md. Also ensure memory scaffolding exists:

• MEMORY.md
• memory/ directory for daily notes
• MEMORY.md must be a standalone workspace-root file, not a section inside AGENTS.md.

Bootstrap Interview Flow

1. Read existing core files if they exist.
2. Ask exactly five bootstrap-priority questions from references/question-bank.md.
3. Synthesize answers into practical defaults.
4. Write or refresh all core files listed above.
5. Set rational session-management and memory defaults:

• session startup read order
• when to create/update daily memory notes
• what belongs in MEMORY.md vs daily files
• heartbeat cadence and quiet-hour behavior

6. Keep BOOTSTRAP.md short, with current setup status and next steps.
7. Summarize what was created, what was updated, and any remaining gaps.

Identity Refresh Flow

1. Read IDENTITY.md.
2. Ask exactly five identity-priority questions from references/question-bank.md.
3. Update IDENTITY.md with:

• ## Backstory
• ## Behavioral Guardrails
• ## Growth Arc

4. Map each update directly to user answers.
5. Summarize changes and open questions.

Writing Constraints

• For IDENTITY.md, preserve existing header fields unless asked to change them.
• Keep sections concise and concrete.
• Explicitly document boundaries/refusals.
• Do not invent personal history if user input is missing; mark assumptions clearly.

Completion Criteria

Finish only when:

• five questions were asked (unless user requested otherwise), and
• required target files for the selected mode were updated with concrete, non-placeholder content.