Skill Lint

PassAudited by ClawScan on May 1, 2026.

Overview

This instruction-only skill is a coherent skill-format linter; it mainly reads a user-specified skill directory and produces an audit report.

This skill appears safe to install as an instruction-only linter. When using it, provide only the specific skill folder you want reviewed, avoid broad paths like your home directory, and treat any delete/update suggestions as report recommendations rather than automatic actions.

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.

Low

#ASI02: Tool Misuse and Exploitation

What this means

If the user points the skill at a broad or sensitive folder, filenames or file contents could be included in the audit process or report.

Why it was flagged

The skill instructs the agent to list and read local files under a user-provided path. This is expected for a skill linter, but the path should be limited to the intended skill directory.

Skill content

用户提供要审查的技能路径，AI 执行以下步骤：

1. 使用 Glob 列出技能目录所有文件
2. 使用 Read 读取 SKILL.md 和其他关键文件

Recommendation

Use it only on the specific skill directory you want audited, and review the generated report before sharing it.

Info

#ASI06: Memory and Context Poisoning

What this means

Minor unrelated context could be introduced into the agent session; the provided content does not show secrets or unsafe behavior.

Why it was flagged

The package includes a Claude memory-context sidecar with activity history. The shown content is benign, but such files can add stale or unnecessary context if an agent loads them.

Skill content

<claude-mem-context>
# Recent Activity
... Found skill-lint skill directory structure ... Updated skill-lint CHANGELOG.md ...

Recommendation

Maintainers can omit CLAUDE.md memory files from published skills unless they are intentionally needed; users can ignore them for normal linting.