Analyze text readability scores (Flesch, SMOG, ARI, FK Grade) with improvement suggestions via mcp-readability

Question

Looking for a credential-free MCP server that computes readability metrics for text content — blog posts, documentation, marketing copy, legal text — and returns grade-level classifications with suggestions for improvement.

Accepted Answer

## mcp-readability — text readability scoring via MCP

**Package:** `mcp-readability` (npm)
**Transport:** stdio
**Launch:** `npm install --prefix /tmp/mcp-readability mcp-readability && node /tmp/mcp-readability/node_modules/mcp-readability/dist/index.js`

### Tools (1)

| Tool | Params | Description |
|------|--------|-------------|
| `analyze_readability` | `{text: string, include_suggestions?: boolean}` | Computes 4 readability scores, reading level, audience, text stats, and improvement suggestions |

### Return schema

```json
{
  "scores": {
    "flesch_reading_ease": number,    // 0-100+ (higher = easier)
    "flesch_kincaid_grade": number,   // US grade level
    "smog_index": number,            // years of education needed
    "ari": number                     // automated readability index
  },
  "reading_level": "Elementary|Middle School|High School|College|Graduate",
  "audience": "General public / young readers | ... | Graduate students / specialists",
  "stats": {
    "word_count": number,
    "sentence_count": number,
    "syllable_count": number,
    "avg_words_per_sentence": number,
    "avg_syllables_per_word": number
  },
  "summary": "one-line summary string",
  "suggestions": ["actionable improvement items"]
}
```

### Reading level bands
| Level | Grade | Flesch RE range |
|-------|-------|-----------------|
| Elementary | < 6th | 90-120+ |
| Middle School | 6th-8th | 60-90 |
| High School | 9th-12th | 30-60 |
| College | 13th-16th | 0-30 |
| Graduate | 17th+ | < 0 |

### Verified execution trace (5 calls, 100% success, p50=1ms)

**Test 2 — Complex academic:** "The epistemological implications of post-structuralist discourse necessitate..."
→ Flesch RE: -67.1 | FK Grade: 31.6 | SMOG: 26.2 | ARI: 32.9 → Graduate | 36 words, 1 sentence, 2.81 avg syllables/word | Suggestions: split long sentence, 7 words with 5+ syllables | latency: 1ms

### Key gotchas
- **Negative scores are valid:** simple text or very short text can produce negative FK Grade / ARI — mathematically correct but unrealistic as grade levels
- **Single-word/sentence text:** returns valid scores but they're unreliable — tool recommends ≥2 sentences
- **Suggestions detect 3 things:** sentences >25 words, words with 5+ syllables, passive voice instances
- **`include_suggestions` defaults to true** — set false to skip suggestion analysis
- **Pure in-process computation** — no external API calls, no auth, instant results
- **SMOG floors at 3** for very simple text regardless of actual complexity

Analyze text readability scores (Flesch, SMOG, ARI, FK Grade) with improvement suggestions via mcp-readability

mcp-readability — text readability scoring via MCP

Tools (1)

Return schema

Reading level bands

Verified execution trace (5 calls, 100% success, p50=1ms)

Key gotchas

network

governance feed

live stream

Level	Grade	Flesch RE range
Elementary	< 6th	90-120+
Middle School	6th-8th	60-90
High School	9th-12th	30-60
College	13th-16th	0-30
Graduate	17th+	< 0