Strip the slop. Keep the insight.

A fast first draft only gets you so far when it still sounds like everyone else's AI-generated copy. Slop Sentry catches the AI tells before your readers do, from the em-dash pileups to "It's not about X. It's about Y." Then it rewrites the draft until it reads like content you can stand behind.

Slop Sentry works inside the tools your writers already use: Claude, ChatGPT, Suits.ai, and any AI assistant that supports MCP, the plug-in standard that lets them call outside tools.

Start free See how it works Understand our philosophy

No credit card. 5,000 words a day on the free plan.

Who it's for

Built for agencies and in-house teams that create content for a living.

Get the efficiency of AI-assisted writing without sacrificing clarity, substance, or voice.

Agencies

Produce work across multiple clients without letting the model flatten them into one generic voice. Slop Sentry removes familiar AI habits, keeps each client recognizable, and brings every draft into AP style compliance.

In-house teams

Use AI across marketing and communications without making every expert sound the same. It keeps experts recognizable and clears vague AI language from everything else your team publishes.

Before and after

The same message, without the AI fingerprints.

Go from confusing, jargon-heavy writing to clear prose with the press of a button.

First draft

6 violations
In today's fast-paced digital landscape, brands need more than content—they need connection. We're thrilled to announce a game-changing partnership with Acme Analytics. Starting in March, its dashboard will seamlessly plug into our reporting platform, helping client teams unlock campaign results in one place instead of exporting spreadsheets from three tools. This isn't just a dashboard. It's a paradigm shift. The result? Early testers were able to delve deeper and cut weekly reporting from four hours to 40 minutes.
  • Empty opener "In today's fast-paced digital landscape"
  • Mirrored contrast "more than content—they need connection"
  • Banned buzzwords "game-changing," "seamlessly"
  • Mirrored logic "This isn't just a dashboard. It's a paradigm shift."
  • Self-asked question "The result?"
  • Banned verbs "delve," "unlock"

Sentry rewrite

CLEAN
We're partnering with Acme Analytics. Starting in March, its dashboard plugs into our reporting platform, so client teams see campaign results in one place instead of exporting spreadsheets from three tools. Early testers cut weekly reporting from four hours to 40 minutes.

The rewrite retains every original fact and removes all six flagged writing patterns.

Add Slop Sentry to your AI assistant and run the same checks on your own draft. Start free.

The approach

The most sophisticated AI writing quality system available.

Most tools stop at pattern matching or a long prompt. Slop Sentry runs a multi-engine editorial pass across the entire skill: exact code where the rule is exact, and specialized judgment where context decides.

400+

Checks across the full skill

More than four hundred checks cover banned language, structure, rhythm, formatting, repetition, substance, tone, and AP Style. One request exercises the whole library, not a short blocklist.

Beyond patterns

Judgment where patterns fail

Code catches exact AI tells. Separate engines evaluate whether the draft says anything concrete, slips into hokey tone, or breaks AP Style. Pattern lists alone cannot answer those questions.

Rewrite loop

Clean means clean

Every revision goes back through the same checks. Fixes that introduce new problems get caught before the draft leaves the conversation.

What it catches

One request runs the whole editorial pass.

Ask your AI assistant to run the Sentry. The Sentry then calls every check below and rewrites until the copy comes back clean. If you would rather see the findings first, the substance, hokiness, and AP Style checks each run on their own, and the full report shows all four verdicts, AI-tell scan included.

The Sentry Primary tool

slop_sentry

One call checks your content for substance, hokiness, AI tells, and AP Style, then rewrites it until every pass reports CLEAN.

1
Substance Catches writing that sounds good but says nothing and contradictions inside the draft.
2
Hokiness Filters try-hard, motivational-poster tone before the draft moves to the deterministic scan.
3
AI tells Scans for 400+ banned words and phrases, plus 30+ additional checks for structure, rhythm, formatting, repetition, and more.
4
AP Style Based on the 2026 AP Stylebook, with 98.2% accuracy across 198 independent test runs.

Every rewrite goes back through the checks, catching new issues introduced while the previous ones were fixed.

Standalone tools

Choose the right tool for the job.

Run a focused review, audit all four checks without changing the draft, or load the workflow instructions for your AI assistant.

Substance check

substance_check

Use it to find empty claims, unclear logic, contradictions, and broken connections. Six independent agents examine the draft; Slop Sentry fixes their flags, then a fresh agent checks the revision.

Included in the Sentry · Phase 1

Hokiness check

hokiness_check

Use it when the draft feels forced, grandiose, or overly earnest. It reads the piece through either a social or professional-audience lens and shows your writing agent which lines to recast.

Included in the Sentry · Phase 2

AI tells guard

guard_sentry

Use it for final cleanup of recognizable model habits. Code checks 400+ banned words and phrases and 30+ writing patterns, rewrites the flagged passages, and checks the draft again.

Included in the Sentry · Phase 3

AP Style check

ap_style_check

Use it to check and revise your draft for compliance with the 2026 AP Stylebook.

Included in the Sentry · Phase 4

The full report

quality_check

Use it when you want to see every finding before anything changes. It runs all four checks and returns one report with no rewrites.

Detection only · All four checks

Agent instructions

get_skill_procedure

Use it to understand which Slop Sentry tools are available, what each one does, and how to use them.

Utility tool

Where the checks come from

Built from thousands of real-world examples.

Slop Sentry comes from ongoing observation and analysis of AI-assisted writing across articles, social posts, emails, proposals, press releases, guides, and more. The tool adapts its review to the format, applying different standards where the writing calls for them, and its checks keep evolving as new patterns emerge.

400+ checks across the full skill: language, structure, substance, tone, and AP Style
10 content types with checks that adapt to the format
Continuously updated as new AI writing patterns emerge

More than a blocklist

Smarter than pattern matching alone.

Slop Sentry combines precise, code-based checks with editorial judgment — over 400 checks that pattern lists and one-shot prompts cannot enforce reliably.

What matters
How Slop Sentry handles it
Prompt-based writing cleaner
Breadth of coverage
Over 400 checks across the entire skill, from banned language and structure to substance, tone, and AP Style.
Usually a short list of banned words or a prompt the model may skip.
Exact checks
Code enforces exact rules for AI tells, rhythm, formatting, and repetition — no guesswork on matches.
Asks the model to interpret and follow a written list of instructions.
Editorial judgment
Specialized engines evaluate substance, hokiness, and AP Style where pattern matching is not enough.
Can describe the desired standard and ask the model to judge its own work.
Voice preservation
Tone checks can account for an active brand voice instead of treating every distinctive phrase as a problem.
Can describe the voice, but the model applies that direction as it rewrites.
Rewrite and recheck
Every edit goes back through the checks until the draft is clean or the rewrite budget is reached.
Usually returns one revision unless the user asks it to inspect and revise the draft again.
Continuous updates
The rule set evolves as new patterns emerge. Everyone connected to Slop Sentry gets the updated checks.
Has to be updated wherever it was copied when the standard changes.
One connection
Claude, ChatGPT and Suits.ai can call the same maintained system after a one-time setup.
Has to be added and maintained separately in each tool or workspace.

Setup

Connect Slop Sentry to the tools you already use.

Two minutes of setup, no code required. Paste one snippet into your AI assistant's settings, restart it, and the tools appear. The guide walks through every step.

View the setup guide 
{
  "mcpServers": {
    "slop-sentry": {
      "url": "https://mcp.slopsentry.ai",
      "headers": {
        "Authorization": "Bearer YOUR-API-KEY"
      }
    }
  }
}
Looking for developer tools? View Python CLI & Agent integration details
$ slopsentry check draft.md --type blog
❌ FAILED - 3 violations.
HYPOPHORA (1): Short rhetorical question...
BUZZWORD (1): Banned AI buzzword: "delve"
AP_STYLE (1): Capitals/numbers compliance...
$ slopsentry rewrite draft.md --provider remote
Job initiated. ID: 4f2a... Polling for results...
Iteration 1 complete. Content is now CLEAN.

CLI companion and agent skills

The same checks in your scripts, pipelines, and editors.

Developers on your team can run the same checks from the command line, inside scripts, publishing pipelines, and text editors. Teams without a developer can paste our SKILL.md rulebook straight into an AI assistant's instructions instead.

  • Agent workflow: Your AI assistant checks the draft, applies the verdicts, and re-checks its own rewrite, following the same steps every time.
  • Automated publishing checks: Stop a document from going out when the scan finds robotic buzzwords or AI tells.

Questions

What you need to know before putting Slop Sentry into your content workflow.

How is Slop Sentry different from the prompts or writing cleaner we already use?

A prompt relies on a model to follow every instruction. The longer your instructions get, the harder it becomes for the model to retain them all.

Slop Sentry runs over 400 checks across the entire skill. Code enforces exact rules for banned language and structural patterns. Separate engines reserve LLM judgment for substance, hokiness, and AP Style — the questions a pattern list cannot answer.

It also keeps going. A prompt typically produces one rewrite, and that rewrite can remove one AI tell while introducing another. Slop Sentry sends the revised draft back through its checks so it can catch new problems as they appear.

Its detector library is maintained and expanded as new AI writing habits emerge. You get the updated checks without rebuilding your prompt every time the models develop another recognizable pattern.

In addition, typical writing cleaners focus on surface-level AI tells. Slop Sentry starts earlier by asking whether the draft says anything meaningful, contradicts itself, or slips into hokey language. There is little value in polishing the wording when the underlying point is empty or unclear.

Will it preserve each client's or expert's voice?

Slop Sentry targets identifiable writing problems instead of imposing one house style. Its rewrites are instructed to preserve the draft's meaning and voice while removing the flagged language. Enterprise plans can also include custom brand voice profiles.

Could it change the facts or meaning of my draft?

The rewrite instructions preserve names, numbers, quotes, facts, and the writer's intended point. Substance Check also looks for contradictions inside the draft. Any AI rewrite still deserves human review, especially when the wording has been legally approved or depends on specialized subject-matter expertise.

How much control do I have over the edits?

You can run the full Sentry and receive a revised draft, call any check on its own, or request a detection-only report that changes nothing. The original draft and the results stay visible in your AI conversation, so you can review, reject, or revise any change.

What kinds of writing can I run through it?

Slop Sentry adapts its checks for articles, blog posts, LinkedIn posts, emails, proposals, presentations, guides, press releases, image copy, and general business writing. Choose the format when you run it so the rules match the way that piece is supposed to read.

What happens to confidential or unpublished drafts?

We never use your content to train models, and text is encrypted in transit. During beta, select execution details, including query text, may be logged temporarily to verify quality and reliability. We plan to eliminate content logging after beta. Do not send material that your policies prohibit third-party AI services from processing.

How accurate are the checks?

The AP Style check operates with 98.2% accuracy across 198 independent test runs. The AI-tell guard combines exact code-based matches with contextual judgment, while Substance and Hokiness evaluate questions that do not have one objectively correct answer. Use the detection-only report when you want to review every finding first.

How does Slop Sentry fit into our existing workflow?

Add it to Claude, ChatGPT, Suits.ai, or another MCP-compatible assistant, then ask that assistant to run Slop Sentry on a draft. The findings and revision come back in the same conversation. Setup does not require a developer; the guide walks through the connection step by step.

Pricing

Choose the level of review your team needs.

Start with AI-tell cleanup for free. Upgrade for the full four-part editorial pass.

Try it

Free

$0

For writers testing the quality checks on occasional drafts.

Words per day
5,000 words / day
Rate limit
3 requests / minute
Max input
500 words per call
  • AI-tell detection and cleanup
  • Instant writing feedback
  • Standard writing formats
Most practical

Teams & creators

Pro

Original price: $99/mo

$49/mo

First year — new users

For active content creators and professional editors checking drafts daily.

Words per day
25,000 words / day
Rate limit
20 requests / minute
Max input
2,500 words per call
  • Unified editorial pipeline checks
  • AP Stylebook compliance checking
  • Substance & Hokiness audit checks
  • Automatic memory for approved words
  • Built-in rewrite assistant

Org-wide

Enterprise

$250+/mo

Team access keys, private usage reports, and custom brand voice profiles, plus dedicated onboarding.

Words per day
150,000 words / day
Rate limit
100+ requests / minute
Max input
10,000 words per call
  • Unified editorial pipeline checks
  • AP Stylebook compliance checking
  • Substance & Hokiness audit checks
  • Team access keys
  • Private team usage reports
  • Custom brand voice profiles