MBTI Agent 人格测评

• Verdict: Pass
• Confidence: High
• Analyzed: Feb 20, 2026, 6:04 AM
• Findings: 0
• Version: 1.0.0

Overview

Files and instructions are coherent with a local/offline MBTI scoring skill; no unusual privileges, network calls, or secret access are requested, though the runtime instructions reference a missing test.py file (inconsistency to be resolved).

This skill appears to do what it says: offline MBTI scoring with a 70-question bank and a local scoring script. Before installing/using: 1) Note SKILL.md refers to test.py but test.py is missing from the package — obtain or inspect that driver (or run scoring.py directly) to ensure expected behavior. 2) Review scoring.py (already included) — it performs only local counting/formatting and has no network calls or secret access. 3) Run the code in a sandbox or isolated environment if you want extra safety. 4) If you intend the agent to answer the 70 questions, ensure the agent's output matches the required JSON schema in questions.md so scoring.py can consume it. If you see any instructions that ask the agent to read unrelated files or export data to external endpoints, do not proceed until those are explained.