Markdown Docs Full-Text Search
PassAudited by ClawScan on May 1, 2026.
Overview
This appears to be a local Markdown search tool; the main things to notice are that it stores indexed document text locally and has a minor package metadata mismatch.
This looks safe for its intended use: local full-text search over Markdown documentation. Before running it, verify the package identity because of the slug/version mismatch, and avoid indexing directories that contain secrets, personal notes, or unrelated project files.
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.
Markdown files in the selected directory may be retained in the generated database and later returned to the agent, including any private notes or sensitive text if the user indexed them.
The skill recursively indexes Markdown files and stores full article content and file paths in a persistent SQLite FTS database.
md_files = list(docs_path.rglob("*.md")) ... INSERT INTO articles (title, source_url, content, file_path, file_hash, indexed_at)Index only intended documentation folders, avoid directories containing secrets or private notes, and delete the generated database when you no longer want the content retained.
The mismatch may make it harder to confirm the package is exactly the one intended, although the included code does not show hidden downloads or unsafe behavior.
This bundled metadata differs from the supplied registry/SKILL identity of md-docs-search version 1.0.2, creating a package identity inconsistency.
"slug": "competitor-docs", "version": "1.0.1"
Verify the publisher/package contents before use; maintainers should align the registry metadata, SKILL frontmatter, and bundled _meta.json.