MdSpliter
PassAudited by VirusTotal on May 11, 2026.
Overview
Type: OpenClaw Skill Name: md-knowledge-spliter Version: 1.0.0 The skill is designed for optimizing large knowledge documents by splitting them into indexed chunks. While it requires file system access (read and write operations on Markdown files), the `assets/evomap_bundle.json` and `assets/gene_knowledge_splitter.json` files explicitly define `forbidden_paths` including `node_modules/`, `.git/`, and `credentials/`, demonstrating an intentional effort to prevent access to sensitive directories. The `SKILL.md` instructions are clear, align with the stated purpose, and contain no prompt injection attempts or directives for malicious actions. There is no evidence of data exfiltration, unauthorized execution, or persistence mechanisms.
Findings (0)
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.
Using the skill may create or update local knowledge chunks and indexes, which could reorganize project documentation.
The asset describes creating directories and Markdown files. This is aligned with the skill's purpose, but it means the agent may modify local knowledge files rather than only read them.
"strategy": ["1. Analyze source file size and structure", "2. Create chunks/ subdirectory in same folder", "3. Split content by topic/section into separate .md files", "4. Create INDEX.md with chunk mapping and tags"]
Confirm the target knowledge directory before allowing file writes, keep a backup of important source Markdown files, and review generated INDEX.md/chunk files.
Future knowledge lookups may depend on the accuracy of the generated index and tags.
The INDEX.md file becomes persistent retrieval context for future answers. If it is stale or incorrect, the agent may repeatedly select the wrong knowledge chunks.
读取 INDEX.md 获取索引 ... 根据查询主题匹配文件 ... 读取对应的分块文件 ... 更新知识时同步更新索引
Review and maintain INDEX.md when knowledge files change, and do not include sensitive private content in reusable knowledge indexes unless intended.