Skill Crafter

PassAudited by ClawScan on May 16, 2026.

Overview

This skill appears to do what it says—help create and register skills—but users should review generated skills before submission because they can affect future agent behavior.

Install/use this if you want an assistant to draft, edit, and register skills. Because the output may become persistent agent behavior, ask to review the final SKILL.md and any generated scripts before registration, especially if the source conversation contained private details or untrusted content.

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

The agent may create local skill files and submit a new or modified skill for registration as part of the requested workflow.

Why it was flagged

The skill instructs the agent to run a local helper script and then submit/register the generated skill. This is expected for a skill-crafting workflow, but it changes the user's skill environment.

Skill content
python3 /sandbox/workspace/skills/skill-crafter/scripts/init_skill.py {name} ... ima_skill_create -d /sandbox/workspace/skills/{name}/
Recommendation

Before registration, ask the agent to show the generated SKILL.md and any scripts if you want to verify what will persist and be submitted.

What this means

Information from the current conversation or later corrections could be written into a reusable skill and affect future agent behavior.

Why it was flagged

The skill can derive workflows from prior conversation and persist lessons into future skill instructions. This is purpose-aligned, but it can accidentally preserve private details or untrusted instructions if not reviewed.

Skill content
从对话历史提取工作流 ... 每次修复一个问题,把经验沉淀到补充说明中
Recommendation

Avoid including secrets or sensitive personal information in skill requirements, and review persisted notes or instructions before publishing.