- Home
- Skills
- Code Quality & Review
- Anti-AI Slop Linter & Vocabulary Scanner
More screenshots
Works with the AI Agents you already use
Anti-AI Slop Linter & Vocabulary Scanner
Audits text, markdown, and PRs for 120+ AI buzzwords, calculates sentence burstiness and lexical diversity, and enforces clean engineering prose.
Free
Anti-AI Slop Linter & Vocabulary Scanner
Example session with this skill installed
Lint my draft PR description in docs/PULL_REQUEST.md for AI buzzwords, measure sentence rhythm, and output recommended engineering replacements.
- Read your context and instructions
- Compiled the anti-ai slop linter
🛡️ ANTI-AI SLOP LINTER — Vocabulary & Burstiness Inspection (v1.0.0)
| Metric | Inspection Result | Benchmark Status |
|---|---|---|
| Analyzed File | docs/PULL_REQUEST.md | 148 Words / 8 Sentences |
| Slop Score | 68 / 100 | 🔴 Critical AI Fluff Detected |
| Burstiness Index | 2.1 std dev (Avg 18.5 words) | 🔴 Monotonous Robot Cadence |
| Lexical Diversity (TTR) | 48.6% | 🟠 Repetitive Vocabulary |
| Buzzword Occurrences | 5 Clichés Flagged | 🔴 Action Required |
🔍 Identified AI Clichés & Suggested Replacements:
- 🔴 Line 1:12 —
"In today's fast-paced realm"
(Context: "In today's fast-paced realm, engineering teams must adapt...")
Suggested: Delete introductory throat-clearing; start directly with technical facts. - 🔴 Line 3:24 —
"delve into"
(Context: "...we delve into the underlying database schema...")
Suggested: Replace withinspect,examine, orprofile. - 🔴 Line 4:41 —
"rich tapestry"
(Context: "...yielding a rich tapestry of microservices...")
Suggested: Replace withdistributed architectureorservice topology. - 🔴 Line 6:15 —
"testament to"
(Context: "This benchmark is a testament to our team's work...")
Suggested: Replace withproves thatordemonstrates. - 🔴 Line 8:05 —
"seamlessly"
(Context: "...allowing clients to seamlessly failover without downtime.")
Suggested: Replace withdirectlyorautomatically.
[CLEANSED PROSE OUTPUT]
Distributed service migration completed. Schema profiling shows 42% latency reduction (p99 dropped from 180ms to 104ms). Automated failover verified across 3 availability zones without packet loss.
Connects securely to your tools. The creator never sees your data.
About this skill
Anti-AI Slop Linter & Vocabulary Scanner
Tired of synthetic, formulaic AI outputs overflowing with corporate fluff, robotic filler words, and monotonous sentence rhythm?
The Anti-AI Slop Linter & Vocabulary Scanner is a lightweight, zero-dependency CLI tool and agent skill designed to audit written content, documentation, pull requests, and agent logs. It flags over 120 known AI clichés, measures lexical diversity, and enforces natural, high-burstiness engineering prose.
⚠️ The Problem: AI Slop Destroys Credibility
When AI models write documentation or pull request descriptions, they exhibit telltale synthetic patterns:
- Pompous Throat-Clearing: Opening essays with "In today's fast-paced world", "Without further ado", or "It is important to remember".
- Repetitive Buzzword Inflation: Overusing "delve", "testament", "rich tapestry", "pivotal", "seamless", "elevate", and "beacon".
- Monotonous Rhythm (Zero Burstiness): Writing sentences of nearly identical 18-word length without dynamic tempo or human cadences.
- Low Lexical Diversity: Repeating identical buzzwords rather than specific domain terms.
⚡ The Solution: Heuristic Linguistic Audit
The Anti-AI Slop Linter inspects text through three deterministic lenses:
- Dictionary Linter (120+ Terms): Pinpoints precise line and column numbers of synthetic buzzwords and provides concrete engineering replacements.
- Burstiness Index (Sentence Variance): Computes sentence length standard deviation. Highlights robotic, uniform phrasing ($\sigma < 3.0$).
- Type-Token Lexical Diversity (TTR): Measures vocabulary richness and flags semantic loops.
- Aggregate Slop Score (0–100): Provides an objective score from clean engineering prose (0–15) to critical synthetic slop (60–100).
🛠️ What's Included:
scripts/slop_linter.py— Fast, zero-dependency Python 3.10+ CLI tool.SKILL.md— Lean Core (<40 lines) agent skill for Claude Code, Cursor, Windsurf, and Antigravity.references/slop_dictionary_and_metrics.md— Exhaustive dictionary of 120+ AI clichés and mathematical metric explanations.
🔒 100% Client-Side Privacy:
Executes completely offline in local memory. No telemetry, no external API calls, zero risk to proprietary documentation or source code.
Requires Python 3. Detection heuristics are specialized for English technical documentation and engineering prose.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 28 days ago
- Free with an Agensi account
Requires
Permissions
Runs 100% offline in local RAM without network egress or API keys.