DEV Community: Promptway

From Invisible to Indexed: The Unglamorous Work of Getting Seen by AI Search

Agnel Nieves — Mon, 15 Jun 2026 20:55:43 +0000

Last week I discovered that my own publication was invisible on its own domain.

Not invisible in the poetic, nobody-reads-my-blog way. Literally invisible. promptway.com, the domain every article on the site declared as its canonical home, was serving an old product prototype I had built months earlier and forgotten about. The actual publication lived on a temporary platform URL. Every AI crawler that found an article was being told "the real copy lives over there," and over there was a different product.

I had written a whole piece about optimizing your site for AI agents and LLMs. The eight layers in that article were all real and all working. And none of it mattered, because the front door had the wrong address on it.

So I spent a day fixing the unglamorous parts. This is the checklist, in plain words. None of it requires you to be a developer, though some steps need one for an hour or two. If part one was about making your site readable to machines, this is about making sure the machines ever show up.

Step 0: check what your domain actually serves

Open a private browser window and type your domain. Not your bookmark. The domain.

Then check three URLs by hand: yourdomain.com/sitemap.xml, yourdomain.com/feed.xml, and yourdomain.com/robots.txt. If any of them 404, that is your whole afternoon right there. Mine did. The publication had all three, but the domain was attached to the wrong project, so the live internet got none of them.

While you are in there, settle the www question. Pick one form of your domain, either yourdomain.com or www.yourdomain.com, and make the other one redirect to it permanently. Search engines treat them as two different sites until you do. I picked the bare domain, made www redirect to it, and also made the hosting platform's free preview URL redirect home, because a copy of your site living on a platform subdomain is a duplicate-content problem you are choosing to keep.

Step 1: let the right robots in

Your robots.txt file is the bouncer at the door. The guest list changed a lot recently, and most sites are still working from an old one.

The crawler that decides whether ChatGPT search cites you is called OAI-SearchBot, and it did not exist when most robots.txt files were written. Same story for Claude-SearchBot, Perplexity-User, and a handful of others. If your file does not mention them, you are relying on default behavior. I would rather be explicit: I allow every major AI search crawler by name, because being cited is the whole point of a publication.

This is a ten-minute edit. Ask whoever manages your site to compare your allowlist against a current guide to AI user agents and add what is missing.

Step 2: put your whole article in the feed

Here is one I had completely wrong. My RSS feed only carried summaries.

A summary-only feed feels safe, like you are protecting the full text. What it actually does is cripple every downstream channel. Feed readers show your readers a teaser and a link. Syndication platforms that import via RSS get nothing worth importing. AI systems that ingest feeds get a paragraph instead of the article.

Full-content feeds are how machines subscribe to you. The fix is technical but small: your RSS feed should carry the complete rendered article, and if you have a JSON Feed, it should too. Your developer will know this as content:encoded. The conversation takes one sentence: "make our feeds full-content."

Step 3: tell the engines when you publish, do not wait to be found

By default, publishing works like this: you post, and then you wait for a crawler to wander by. That can take days.

Two free standards flip it to a push model. IndexNow is a single notification you send when a page is new or updated, and one submission covers Bing, Yandex, and a few other engines at once. Bing matters more than its search share suggests, because most of the web results ChatGPT cites overlap heavily with Bing's top results. WebSub does the same job for feed readers: it pings a hub when your feed changes, and subscribers update within seconds instead of whenever they next poll.

Both are set-and-forget. We wired ours into a small script that runs automatically on publish, and I never think about it. If you publish through a platform like Ghost or WordPress, there is a decent chance a plugin or setting already does this. Turn it on.

Step 4: put a human name on AI-assisted work

Some of the writing on Promptway is drafted by AI personas. That is disclosed openly, profile pages and all. But disclosure alone is not enough anymore, and the sites that got hammered by Google's scaled-content crackdowns had one thing in common: nobody human was accountable for the words.

So every persona-written article on the site now carries a visible line: "Reviewed by Agnel Nieves," linking to my profile. It is the same pattern medical sites have used for years with "medically reviewed by Dr. X." The page also says it in structured data, where I am listed as the editor, and my profile is connected to my real accounts elsewhere so machines can verify I am a person who exists.

One honest note: this is an accountability signal, not a ranking cheat code. It tells readers and crawlers that a named human stands behind the work, which happens to be true. If it were not true, I would fix that first.

Step 5: never ship a blank share card

Paste one of your article links into Slack or iMessage. If the preview is a gray rectangle, you are losing clicks you already earned.

Most articles on my site have generated hero art, but a few do not, and those shipped with no preview image at all. The fix was a small template that auto-generates a clean, typographic card with the article title for any post without art. Every link now unfurls into something. Most platforms have this built in; if yours is custom, it is an afternoon of work for a developer, once, forever.

Step 6: measure the AI traffic separately

Here is the stat that convinced me to finally set up analytics: visitors arriving from AI tools like ChatGPT and Perplexity are a tiny slice of traffic, around one percent industry-wide, but they convert far better than organic search visitors. Someone who clicks through from an AI answer chose your site after the machine already summarized you. That is a warm lead, not a drive-by.

If you lump that traffic in with everything else, you will never see it. The setup is one custom channel group in Google Analytics that buckets referrals from chatgpt.com, perplexity.ai, claude.ai, and gemini.google.com. Ten minutes in the admin panel, and from then on you can answer "is any of this AI optimization work doing anything?" with a number.

The part where I admit a fleet of agents did the typing

Full transparency: I did not hand-write most of this. I described the plan, and a team of AI coding agents executed it, each sized to its task. Small models refreshed the robots file and the redirects. Bigger ones rebuilt the feeds and wrote the publish-time automation. I reviewed the output, caught a few real mistakes (one agent excluded the exact files we needed redirected, another drew a logo with the text colliding), fixed them, and shipped.

That division of labor felt right. The judgment calls, what to build, what to skip, what counted as done, stayed with me. The typing did not have to.

The whole checklist, one more time

Type your domain into a private window and make sure it serves the thing you want found, on one canonical host.
Check sitemap.xml, feed.xml, and robots.txt by hand. Fix the 404s.
Update your robots.txt allowlist for the 2026 crawler landscape.
Make your feeds full-content.
Wire up IndexNow and WebSub so publishing pushes instead of waiting to be pulled.
Put a named, linkable human reviewer on anything AI-assisted.
Generate a fallback share image so no link ever unfurls blank.
Build an AI-referrals channel group in your analytics.

None of this is clever. That is the point. The clever work, the writing, the structured data, the llms.txt file, only pays off after the plumbing is sound. Check the front door first.

The De-Slop Prompt Stack: Six Prompts That Stop Claude and ChatGPT From Sounding Like a Robot

Agnel Nieves — Mon, 15 Jun 2026 20:55:42 +0000

Merriam-Webster's 2025 word of the year wasn't a tech term that crossed over. It was an insult. Slop: roughly, low-quality digital content churned out in volume by AI.

When the dictionary names the thing your content pipeline produces by default, that's worth twenty minutes of your attention. Here are the twenty minutes.

The trust penalty is measurable now

Canva's State of Marketing and AI 2026 report, a Harris Poll of 1,415 marketing leaders and 3,547 consumers across seven countries released in May, found that 97% of marketing leaders now use AI daily. The same report found that 78% of consumers would rather see ads made by people, that 87% say the best advertising still requires a human touch, and that mentions of "AI slop" were up ninefold.

Bynder's Human Touch survey of 2,000 UK and US consumers adds the twist. 52% said they get less engaged the moment they suspect copy is AI-written. But 56% actually preferred an unlabeled AI-written article over the human copywriter's version.

Read those two Bynder numbers together. People can't reliably detect good AI writing. They reliably punish detectable AI writing. The penalty isn't for using the tool. It's for the tells.

So the job is simple to state: remove the tells.

The tells, so you can delete them on sight

Before the prompts, the symptom list. Machine-default writing is recognizable because it always fails the same ways:

Metronome rhythm. Every sentence runs 15 to 20 words, forever.
The connector parade: furthermore, moreover, additionally, consequently.
Giveaway vocabulary: delve, tapestry, landscape, elevate, crucial, robust, journey.
"It's not just X. It's Y." Fine once. A tell when it's every third paragraph.
Threes. Everything arrives in threes.
Confident claims with no source attached.
Em dashes every other sentence. We ban them at Promptway entirely, and the models adore them.

No single item proves anything. Stacked together, they're the smell. The six prompts below kill them at the source.

The stack

The first three are setup prompts. Run them before drafting, or better, save them permanently (instructions at the end). The last three are editing passes you run on a finished draft.

1. The rhythm and banned-words prompt

The two loudest tells are vocabulary and rhythm, so handle both in one standing instruction:

# The rhythm and banned-words rules
Apply these writing rules to everything in this conversation:

Never use these words: delve, tapestry, landscape, elevate, crucial,
robust, seamless, journey, realm, leverage, unlock, harness,
game-changing, transformative.

Never open with "In today's" or "In the world of."

No "furthermore," "moreover," "additionally," or "consequently."
Connect ideas the way people talk: "and," "but," "so," or just start
the next sentence.

No em dashes anywhere. Use a period, a comma, or parentheses.

Vary sentence length on purpose. Some sentences under six words.
Some over twenty. If three sentences in a row share the same
structure, rewrite the third one.

Maximum one exclamation point per piece. Prefer zero.

Why it works: models respond far better to explicit negatives than to "sound more human," which is too vague to act on. What goes wrong: in long conversations the rules drift. Repaste them, or save them permanently so you never have to.

2. The voice transplant

The default model voice is generic-competent, which means it belongs to nobody. Give it yours:

# The voice transplant
Below are samples of my actual writing, about 400 words total.
Study how long my sentences run, which words I repeat, how I open
and close paragraphs, and how often I use contractions.

[PASTE 300 TO 500 WORDS OF YOUR OWN WRITING]

Write the piece I describe next in that voice. Do not reuse my
sentences or my topics. Match the rhythm and the vocabulary level.
If you are not sure whether I would say a phrase, leave it out.

Use writing you produced without trying to impress anyone. Emails to colleagues work better than your polished blog posts. What goes wrong: a sample under 200 words produces a caricature instead of a voice. And the last line matters, because "leave it out" stops the model from filling gaps with its defaults.

3. The coffee test

Our writing guide's oldest rule is to read every draft out loud. This prompt builds the rule into the machine:

# The coffee test
Rewrite this draft as if you were explaining it to a smart friend
over coffee. Contractions are fine. Sentence fragments are fine.
Keep every fact. Cut every sentence that sounds like a press
release. The test: if a sentence would sound stiff read out loud,
rewrite it until it doesn't.

This one earns its spot on tone alone. It converts announcements into explanations, which is most of what "sounding human" means.

4. The specificity pass

Vagueness is slop's natural register, and it's also where AI writing quietly lies. This pass fixes both:

# The specificity pass
Go through this draft and replace every vague claim with a concrete
one. "Saves time" becomes a number of minutes or hours. "Many
companies" becomes a count, a sourced percentage, or one named
example. "Recently" becomes a month and a year.

Important: if you do not have the real specific, do not invent one.
Write [NEED SPECIFIC] in its place so I can fill it in myself.

The [NEED SPECIFIC] line is the load-bearing part. Demand specifics without it and the model will manufacture them. A marked gap is useful. An invented number is a liability with your name on it.

5. The honest objection

Human writing argues with itself a little. Default AI writing never does, which is part of why it reads like a brochure:

# The honest objection
Find the weakest claim in this draft. At roughly the halfway point,
add the strongest objection a skeptical reader would raise against
it, then answer the objection honestly in two or three sentences.
If the objection partly wins, concede that part. Do not strawman
the reader.

A conceded point reads more human, and more trustworthy, than a parade of wins. It also forces you to know which of your claims is the weak one, which is worth knowing before your readers find it for you.

6. The true-anecdote opener

Nothing reads more human than a real, specific moment. Nothing reads more fake than an invented one, so this prompt comes with a fence:

# The true-anecdote opener
Here is something that actually happened: [DESCRIBE THE REAL MOMENT
IN TWO OR THREE SENTENCES: who, where, what went wrong or what
surprised you].

Open the piece with a short scene built only from those details.
Do not add details I did not give you. Keep the scene under 80
words, then move into the main point.

Ask a model for "an engaging opening story" and you get confident fiction. Feed it a real moment and restrict it to that moment, and you get the one thing slop can never fake: something that happened.

Save it once, stop repasting

Don't run this by hand every morning. Prompts 1 through 3 belong in your standing instructions: custom instructions or a Project in ChatGPT, and your project's custom instructions in Claude. Every new chat then starts with your rules and your voice already loaded. Prompts 4 through 6 work best as an editing pass you run on finished drafts before anything ships.

This is the same logic as putting your constraints at the top of the prompt: the rules you cannot afford to lose should be the first thing the model reads, every time, without you having to remember them.

Total setup time is about fifteen minutes, once.

What these prompts will not do

They fix tone. They do not fix truth.

A model can produce a beautifully rhythmic, perfectly voice-matched paragraph built around a wrong number. Verify every figure, name, date, and quote before you ship, every time. Slop that's also wrong is worse than slop.

And keep humans on the pieces where the reader needs to feel a person on the other end: the apology, the crisis note, the founder letter. The Canva data says people don't hate AI content as much as they want to feel that a human took responsibility for it. On some pieces, that feeling is the content.

The point was never the detector

AI detectors are coin flips anyway, so fooling them was never a goal worth having. The goal is to stop wasting your reader's attention on writing that sounds like everyone else's default settings.

Run the stack, then read the draft out loud. If it sounds like a person, ship it.