Skillv0.0.1

ClawScan security

个人发展复盘助手 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 5, 2026, 9:54 AM

Verdict: Benign
Confidence: medium
Model: gpt-5-mini
Summary: The skill's requirements and instructions match its stated purpose (generating a personal review from text or local files); no installs or credentials are requested, but the runtime instructions allow reading arbitrary local files so users should avoid supplying sensitive paths.
Guidance: This skill appears coherent for producing structured Chinese personal-review reports. It will read local files if you supply file paths — do not give paths to sensitive system files (ssh keys, password stores, configs). Prefer pasting or uploading the specific diary/text you want analyzed rather than pointing to broad directories. The SKILL.md states “All data stays local / No network calls,” but that is an instruction, not an enforced guarantee; consider the platform's data-handling policy before sending highly sensitive content.

Purpose & Capability: okThe name/description (生成个人复盘报告) align with the declared tools and behavior: it needs user text or local files and uses read_file/read_dir/date_utils to produce structured reports. No unrelated binaries, env vars, or installs are requested.
Instruction Scope: noteSKILL.md stays within the task (summarize diaries/files into sections). It explicitly instructs use of read_file/read_dir for local files and to format OS-specific Bash examples if a file link is given. This is reasonable for file-based input, but the instructions permit reading any local path supplied by the user — there is no built-in limit on which files may be accessed.
Install Mechanism: okInstruction-only skill with no install spec and no code files — lowest-risk install footprint.
Credentials: okNo environment variables, credentials, or config paths are requested. Declared toolset matches the skill's needs.
Persistence & Privilege: okalways is false; autonomous invocation is allowed (the platform default). The skill does not request persistent or elevated privileges or modify other skills' settings.