Article2Book
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This skill appears purpose-aligned and benign, but it works by reading and summarizing files in a folder you choose, so keep that folder limited to content you are comfortable having analyzed and saved locally.
This skill is reasonable to use for reorganizing existing writing or course materials. Before running it, place only the intended materials in a dedicated folder, avoid including confidential client or personal files unless you are comfortable with them being read and summarized, and review the generated output folder for retained notes or inventories.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
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.
If pointed at a broad folder, the skill may index many local documents and create summaries/previews in output files.
The helper script recursively indexes supported files in the user-provided source directory and writes local inventory outputs. This is expected for the skill's content-planning purpose, but users should scope the directory carefully.
candidate_files = sorted(... for path in source_dir.rglob("*") ...); write_csv(output_dir / "article_inventory.csv", records)Use a dedicated source folder containing only the materials you want analyzed, and review or delete generated inventory files if they are no longer needed.
Sensitive or confidential material in the selected folder may be reflected in local planning notes, inventories, or generated drafts.
The skill directs the agent to read all candidate text materials and optionally persist reading notes. This is central to the purpose, but it can create durable summaries of private source content.
Agent 应尽量全部通读... 通读记录建议沉淀到：processed/agent_reading_notes.md
Redact or exclude confidential files before use, and check the `书稿策划输出/` and `processed/` outputs for sensitive summaries.
For highly confidential materials, additional agent contexts may increase the amount of processing surface.
The workflow allows subagents to read batches of source materials. This is purpose-aligned and bounded by batch assignment, but it means document content may be shared across auxiliary agent contexts.
当素材很多时，可以使用 subagent 分批并行通读；但最终的母题判断... 必须由主 Agent 统一收束。
For sensitive projects, ask to avoid subagents or use a smaller, pre-redacted source folder.