Skip to content

text

String manipulation and content extraction. Pure compute — auto-registered by the installer.

Usage

bash
native-mcp text

No flags.

Tools

slug

URL-safe slug from any string. Lowercase, non-alphanumerics collapse to -, leading/trailing dashes trimmed. Unicode letters and digits are preserved.

FieldTypeRequiredDescription
inputstringyes

regex_match

Apply a Go RE2 regex.

FieldTypeRequiredDescription
patternstringyesRE2 syntax.
inputstringyes

Returns the first match plus capture groups, or no match.

regex_replace

FieldTypeRequiredDescription
patternstringyesRE2 syntax.
inputstringyes
replacementstringyes$1, $2, ${name} reference capture groups.

diff

LCS-based line diff with unified-style markers (, -, +).

FieldTypeRequiredDescription
astringyesOriginal.
bstringyesUpdated.

markdown_extract

Strip markdown formatting and return readable plain text. Removes headers, emphasis, links, code spans/blocks, list markers, blockquotes, horizontal rules.

html_extract

Strip HTML tags and return readable plain text. Drops <script>, <style>, <nav>, <header>, <footer>, <aside>, <noscript>, <head> blocks. Decodes common HTML entities.

Pair this with fetch.fetch_url to turn a webpage into clean text for the model.

Register in Claude Code

bash
claude mcp add native-text native-mcp text

Released under the MIT License.