Skillv1.0.0

ClawScan security

Mbti Persona · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 6, 2026, 2:15 PM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill does what it claims: a local MBTI persona configurator that reads/writes a JSON file under ~/.openclaw and has no external network access or extra credentials.
Guidance: This skill is locally scoped and appears safe: it simply creates/reads ~/.openclaw/mbti-config.json with the chosen MBTI profile and prints the profile to stdout. Before installing, if you care about provenance, note the source is unknown (no homepage) — you may want to inspect the shipped script (scripts/mbti_config.py) yourself. Also be aware the config is stored in plaintext in your home directory; if that matters, restrict filesystem permissions or store it elsewhere.

Purpose & Capability: okName and description match the included pieces: a SKILL.md that describes MBTI personas and a small Python script that sets/gets/lists an MBTI type. No unrelated credentials, binaries, or services are requested.
Instruction Scope: okSKILL.md only instructs the agent to run the included script (set/get/list) and documents the config path (~/.openclaw/mbti-config.json). The instructions don't ask the agent to read unrelated files, environment variables, or transmit data externally.
Install Mechanism: okNo install spec — instruction-only plus a shipped Python script. The script is plain Python with no dependencies, no downloads, and no archive extraction.
Credentials: okThe skill requires no environment variables or external credentials. It writes a small JSON config to the user's home directory (~/.openclaw/mbti-config.json), which is proportional to the stated feature.
Persistence & Privilege: okalways is false and the skill does not modify other skills or global agent configuration; it only creates/updates its own config file in ~/.openclaw. The agent-autonomy default is unchanged.