wsc

Wsc MCP Server

WSC helps writers and developers inspect prose through AI assistants for style issues and research-cited patterns associated with AI-generated text. It reports contextual findings, returns cleaned text for repeated words, and supports local file checks.

Server facts

wsc MCP Server

Tech

Runtime / language
node
Transports
Not specified
Remote available
No
Authentication
Not specified

Activity

GitHub stars
9
Last commit
Data as of

Tags

  • Documentation
  • ai detection
  • ai slop
  • github action
  • linter
  • mcp
  • mcp server
  • prose linter
  • technical writing
  • writing tools

Overview

Prose review in WSC centers on specific matches rather than an authorship verdict. Its check_text tool examines conversational text for vague wording, passive constructions, repeated words, lengthy sentences, nominalizations, hedges, filler adverbs, and patterns associated with AI-generated writing. Reports identify line and column positions, show the matching passage with context, and explain AI-tell findings using cited research. These flags do not establish who wrote a passage. Writers can review findings through an assistant, while developers can incorporate checks into workflows using the project's API, CLI, or GitHub Action.

For documents on disk, the local-only check_file tool reads without modifying the source. Markdown analysis excludes code blocks and tables, and configuration is discovered in the containing directory or its parents unless supplied explicitly. The fix_duplicates tool returns revised text and a record of removed repetitions without writing changes; other findings remain report-only. To inspect detector vocabulary before adjusting settings, list_word_lists provides counts, configuration keys, examples, and access to the browsable library. MCP access is available through a hosted Streamable HTTP connection or a local stdio server.

Related MCP servers