dislop
OfficialThis server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
dislop is an open-source, cross-modal quality gate that inspects prose, web interfaces, slide decks, and office documents before delivery. Instead of feeding drafts into an opaque "AI humanizer" that introduces grammar errors or alters facts, dislop runs deterministic probes first, pairs every flagged pattern with concrete before/after editorial exemplars, and keeps domain taste modular through pluggable Taste Packs.
Why dislop Exists
First-generation "anti-AI" rewriters fail in real production work for three reasons:
Blind full-text rewriting degrades facts: Unconstrained LLM rewrites frequently mutate numbers, units, proper nouns, and citations to dodge statistical detectors.
Text-only tools miss visual and structural tells: A deck with 260 words crammed onto Slide 4, a landing page wrapped in default indigo-purple-pink gradients, or a PDF leaking local filesystem paths fails reader trust before anyone reads paragraph two.
Good taste is domain-specific, not universal: What makes a client research brief crisp is different from what gives a 6,000-word literary essay breathing room.
dislop separates deterministic inspection (src/dislop/, bin/) from curated editorial standards (packs/), exposing both a fast local CLI and a stdio Model Context Protocol (MCP) server for coding and writing agents.
Related MCP server: slop-guard
Quick Start
Node.js / npx (Zero External Dependencies)
# Audit a Markdown file, HTML page, or entire repository
node ./bin/dislop.mjs check README.md
node ./bin/dislop.mjs check ./docs --pack business-brief
# List installed Taste Packs or inspect a rule's editorial rationale
node ./bin/dislop.mjs packs
node ./bin/dislop.mjs explain zh-cliche-connectorPython / uv (Includes Zero-Dependency .docx & .pptx Inspection)
# Run directly from source or install into a virtual environment
PYTHONPATH=src python3 -m dislop.cli check ./README.md
PYTHONPATH=src python3 -m dislop.cli check ./proposal.pptx --pack deck-and-uiConnect as an MCP Server (Claude Code, Codex, Cursor, Antigravity)
Add dislop to your MCP client configuration so your agent checks drafts and pulls editorial exemplars before handing files back to you:
{
"mcpServers": {
"dislop": {
"command": "node",
"args": ["/path/to/dislop/bin/dislop-mcp.mjs"]
}
}
}Exposed MCP tools:
dislop_check: Audit raw text, a single file (.md,.html,.svg,.docx,.pptx), or a directory against a Taste Pack.dislop_packs: List installed Taste Packs, supported modalities, and rule counts.dislop_exemplars: Retrieve annotatedbad -> goodrewrites and editorial rationales so the calling agent can revise flagged passages without corrupting surrounding facts.
Built-in Taste Packs
Pack ID | Target Modality | What It Catches |
| Proposals, memos, research reports ( | Filler transitions, inflated verbs, local path leaks, excessive em-dash cadence |
| Long-form essays, columns, non-fiction ( | Moralizing paragraph endings, forced triadic symmetry, translationese passive scaffolds |
| Slide decks & web UI ( | Per-slide word walls ( |
Dogfooding Standard
Every Markdown document, SVG asset, and release note in this repository is checked by dislop itself in CI (.github/workflows/ci.yml). Pull requests that trigger slop findings in repository docs fail the build until rewritten cleanly.
Repository Layout
dislop/
├── assets/ # Brand mark (logo.svg) and masthead banner (banner.svg)
├── bin/ # Zero-dependency Node.js CLI & stdio MCP server
├── src/dislop/ # Zero-dependency Python CLI, OOXML (.docx/.pptx) inspector & MCP server
├── packs/ # Pluggable Taste Packs (rules + before/after exemplars)
├── docs/ # English & Chinese architecture specifications
└── tests/ # Unit, OOXML fixture, and self-dogfooding test suiteLicense
Licensed under the Apache License 2.0.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
DraftCheck: flags AI-writing patterns in prose with fix hints. Deterministic linter, not a detector.
30+ deterministic dev and delivery tools: linting, metrics, diagrams, model advice, AI-text checks.
Scan, fix and justify your application code until it is approved to ship.
Scan AI agent skills, MCP server packages and MCP configs before you install them. Static analysis.
Related MCP Servers
- FlicenseNot gradedqualityAmaintenanceEnables agents to analyze text strings or files for formulaic AI writing patterns, returning a 0–100 score, band label, specific violations with context, and concrete advice. It performs this rule-based linting locally without LLM judges or API calls.-
- FlicenseNot gradedqualityAmaintenanceEnables agents to analyze inline prose or files on disk for formulaic AI writing patterns, returning a 0-100 score, band label, specific violations with context, and concrete advice. Uses purely programmatic, configurable rules with no LLM judge or API calls.-
- FlicenseNot gradedqualityAmaintenanceEnables agents to analyze text or files and receive numeric slop scores, band labels, specific violations with context, and concrete writing advice, using rule-based diagnostics with no LLM judge or API calls.-
- FlicenseNot gradedqualityAmaintenanceEnables agents to score prose from 0–100 for formulaic AI writing patterns using rule-based linting, returning diagnostics such as band labels, violations with context, and advice. It can analyze inline text or files without API calls or LLM judges.-