SkillTree

ReviewAudited by ClawScan on May 10, 2026.

Overview

SkillTree is mostly a gamified personalization skill, but its efficiency mode may encourage the agent to act before asking on emails, files, calendars, or messages.

Before installing, decide whether you want this kind of persistent, chat-history-based personalization. If you use it, avoid enabling any “do first, ask later” behavior for emails, messages, file changes, calendar updates, purchases, public posts, or other actions you would want to approve first.

Findings (2)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

If the agent has access to email, calendars, files, or messaging tools, it may take actions before asking you, causing unwanted sends, edits, or replies.

Why it was flagged

This explicitly encourages replacing confirmation with immediate execution for an external communication action. Email sending and similar actions may be irreversible or high-impact, and the skill does not define a clear safety boundary for when confirmation is still required.

Skill content
Decision Executor
```
Before: "Do you want me to send this email now?"
After: "Email sent ✓ [Undo]"
```
Recommendation

Keep explicit confirmation for sending messages or emails, changing files, modifying calendars, posting publicly, purchases, or other irreversible actions. Treat “do first, ask later” as only suitable for low-risk tasks.

What this means

The agent may remember preferences or personal details from prior conversations and use them later; incorrect or sensitive memories could affect future replies.

Why it was flagged

The skill is designed to analyze recent chat history and maintain profile/snapshot state. This is purpose-aligned, but it means personal chat context can be reused to shape future behavior.

Skill content
1. 分析对话历史 (最近 50 条) ... `evolution/profile.json` ... `evolution/snapshots.json`
Recommendation

Use the skill only if you are comfortable with persistent personalization. Review, reset, or roll back the profile if it stores sensitive, incorrect, or unwanted assumptions.