← Back to Skills Marketplace
248
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install second-me
Description
Build your digital twin through conversation. Your AI learns who you are by talking to you — not by reading a form.
Usage Guidance
This skill appears to do what it says: it will interview a user and persist a personal identity document and conversation logs under ~/.openclaw/data/second-me. Before installing, consider: 1) Sensitive data: SOUL.md and session logs may include private workplace/conflict/personal details — review, restrict, or encrypt that folder and avoid putting credentials or PII into the file. 2) Confirmation channel: the skill references confirming actions with the physical self but doesn't define how — decide and enforce a clear verification method (e.g., an out-of-band confirmation step) if you plan to allow any autonomous actions. 3) Access control: check who/which processes can read ~/.openclaw/data/second-me and consider disk encryption/backups retention policies. 4) Review content regularly and delete or redact records you do not want persisted. If you are uncomfortable with any autonomous capability, keep the skill user-invocable only or disable autonomous invocation at the agent level.
Capability Analysis
Type: OpenClaw Skill
Name: second-me
Version: 1.0.1
The 'second-me' skill is a digital twin builder that conducts a structured interview to create a personal identity document (SOUL.md) stored locally in the user's home directory. While it collects highly sensitive information, including workplace dynamics, professional history, and personal values, this behavior is explicitly aligned with its stated purpose. The instructions in SKILL.md include comprehensive safety guardrails that forbid the agent from exfiltrating data, sharing credentials, or performing unauthorized financial/legal actions, and it lacks any indicators of malicious intent or hidden execution logic.
Capability Assessment
Purpose & Capability
The skill claims to build a digital twin via conversation and all declared behavior is instruction-only: reading/writing a SOUL.md and logging sessions under ~/.openclaw/data/second-me, which is coherent for this purpose. No unrelated binaries, env vars, or external credentials are requested.
Instruction Scope
SKILL.md instructs the agent to run structured onboarding conversations, read an existing SOUL.md, write/append SOUL.md after each topic, and persist conversation logs. This stays within the stated purpose, but the instructions cause accumulation of highly sensitive personal and workplace data (manager relations, politics, personal values). The instructions do not specify how confirmations with the physical user must be performed (channel, verification), which leaves ambiguity about safe decision-making and confirmation flow.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest-risk install profile. Nothing is downloaded or executed on disk beyond the agent writing its own data files as described in SKILL.md.
Credentials
The skill requests no environment variables, credentials, or config paths beyond its own data directory under the user's home. That is proportionate to its stated function of storing a digital identity document and conversation logs.
Persistence & Privilege
always:false (normal). The skill writes persistent files (SOUL.md, progress.json, session logs) to ~/.openclaw/data/second-me which is expected. However, the documentation includes a 'Body & Mind Partnership' section describing how the twin should operate when the human is unavailable; combined with the platform's normal ability for agents to invoke skills autonomously, this creates a risk surface if the confirmation/authorization channel is not strictly enforced. The skill itself includes safety clauses forbidding financial/legal impersonation, but those are high-level and do not define technical enforcement.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install second-me - After installation, invoke the skill by name or use
/second-me - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- File storage paths have been updated to use the `.openclaw/data/second-me/` directory instead of `.claude/second-me/`.
- All references to `~/.claude/second-me/` and its subfolders are now `~/.openclaw/data/second-me/`.
- Onboarding and evolve process logic, conversation topics, and SOUL.md/spec structure remain unchanged.
- No other functional or behavioral changes were introduced.
v1.0.0
Initial release of second-me — a conversational digital twin builder.
- Enables users to build a living digital twin (SOUL.md) through natural conversation, not forms.
- Structured onboarding covers 7 key topics (basics, professional background, work style, workplace dynamics, ambitions, life outside work, values & principles).
- Each topic results in targeted updates to SOUL.md using specific, second-person language (never inventing details).
- Progress is saved and tracked in progress.json, allowing seamless session pick-up and gap detection.
- Evolve Mode supports ongoing updates, contradiction checks, and periodic reviews to keep the digital twin accurate and relevant.
- Built-in, non-negotiable safety & boundary rules ensure secure, ethical twin operation.
Metadata
Frequently Asked Questions
What is Second Me?
Build your digital twin through conversation. Your AI learns who you are by talking to you — not by reading a form. It is an AI Agent Skill for Claude Code / OpenClaw, with 248 downloads so far.
How do I install Second Me?
Run "/install second-me" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Second Me free?
Yes, Second Me is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Second Me support?
Second Me is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Second Me?
It is built and maintained by Paul Leo (@paul-leo); the current version is v1.0.1.
More Skills