Back to skill
Skillv0.1.0

ClawScan security

Weclone Init Twin · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 12, 2026, 2:50 PM
Verdict
benign
Confidence
high
Model
gpt-5-mini
Summary
The skill's code, templates, and runtime instructions match its stated purpose of scaffolding persona markdown files and do not request unrelated credentials or perform network/I/O beyond writing those files.
Guidance
This skill appears to do what it claims: scaffold persona markdown files using bundled templates and a small Python script. Before installing/running: (1) Inspect the included templates and script if you want to confirm content; (2) be aware the default output directory is repo-root/ai_twin (the script computes a repo-root path) and the script will create directories and files there; (3) files are protected from accidental overwrite unless you pass --force, so only use --force when you explicitly want to replace existing persona files; (4) run the script with a safe test output directory first if you want to verify behavior; (5) no network calls or secret access are present, but because the skill can be invoked by the agent, consider restricting autonomous invocation if you do not want the agent to create files without manual approval.

Review Dimensions

Purpose & Capability
okName/description match the included assets and script: templates for profile/state/guardrails and a Python script to render them. Required binary (python3) is appropriate and nothing unrelated (no cloud credentials, no unrelated binaries) is requested.
Instruction Scope
okSKILL.md limits actions to creating or refreshing persona markdown files and requires explicit approval to overwrite. The included script reads bundled templates and writes files to a target directory; it enforces non-overwrite unless --force is passed. Instructions do not reference other system files, secrets, or external endpoints.
Install Mechanism
okNo install spec is provided (instruction-only), and the skill ships only a small Python script plus markdown templates. Nothing is downloaded or extracted during install.
Credentials
okThe skill requires no environment variables or credentials. The only runtime dependency is python3, which is proportional to running the included script.
Persistence & Privilege
notealways:false (normal). The skill will write files to a target directory (default repo-root/ai_twin) and can write to any path the user supplies; this is expected behavior but means users should confirm target paths before running. The agent's autonomous invocation is allowed by default (not unusual) — if you want to prevent automatic creation of files, keep the skill user-invocable only or disable autonomous invocation in agent settings.