Point it at your story. Get six SEO headlines, nine social posts, and five email subject lines back. Maybe they're useful.
Go to navybook.com/D1/seo/, paste your article, and generate headlines and social-media posts. Works in any browser, no installation needed. Good for occasional use or if you're on a machine without the extension.
Drag the bookmarklet to your browser toolbar. When you're editing a post in Athena, one click grabs the article text and sends it to Headline Lab — no copying, no switching tabs, no pasting.
The most integrated option. The extension lives inside Athena's post editor — a Headline Lab button appears next to the Title field. Click it to open a slide-in panel, generate headlines and subheds, and paste them directly into the post with one click. No tab-switching required.
Clicks from Google are down, so I (that is, D1 editor Brad Peniston) asked Claude to whip up a tool that could help us with SEO without exposing/giving away our prepublication work. Which it did, in just a few minutes. Then I asked it to add social post generation. Then I asked it to write almost all of the text on this page.
Now the tool lives on my private domain and it swears it's not sending anything that might be captured and stored.
Please let me know what you think, and how this might be made more useful and easier to use.
Use the webpage, the bookmarklet, or the Chrome extension to generate six SEO-friendly heds and subheds, nine social-media posts tuned for Facebook, X, and LinkedIn, or five email alert subject lines with preview text.
For headlines, you can specify a target keyword and a tone — straight news, urgent/breaking, analytical, conversational, or authoritative. The tool applies SEO rules automatically: front-loaded keywords, 50–60 character targets, active voice. Formats like listicles or question-framing are only used when the article actually supports them — accuracy beats structure.
For social posts, the tool writes to each platform's conventions: punchy for X, conversational for Facebook, analytical and professional for LinkedIn. Both headline and social generation start with a fact-extraction step, so suggestions stay grounded in what the article actually says rather than drifting into inference or hype.
For email subjects, the tool follows inbox best practices: 40–50 character subject lines (fully visible on mobile without truncation), no question-form subjects (they underperform for news alerts), specificity over vague intrigue, and paired preview text that extends — rather than repeats — the subject line.
This tool uses Anthropic's API — a different product from the public Claude.ai chatbot. Under Anthropic's API terms, content sent through the API is not used to train AI models and is not retained after your request is processed. Your unpublished article text is never stored on our server either. Think of it like a fax: the text travels to get processed, the answer comes back, and nothing is kept.
The tool reads your full text to extract key facts before writing. A longer paste means more accurate fact extraction, better keyword identification, and suggestions that actually reflect what your story says.
If you've already done keyword research or know what readers are searching, enter it. The tool will prioritize fitting that phrase naturally into at least some options.
Breaking news needs urgency; a policy explainer needs authority; a features piece might work conversationally. Selecting the right tone produces more usable suggestions.
Each generation is independent. If the first batch feels off — wrong angle, wrong tone — hit Generate Heds again. The fact-extraction step re-runs too, so you may get a different set of grounded options.
Hybrid headlines often outperform pure AI output. Take the keyword placement from one suggestion, the structure from another, and add your own editorial voice.
The 50–60 character window is not a hard rule, but headlines in this range are statistically less likely to be truncated in Google Search and news aggregators.
X posts over 280 characters can't be published as-is. If a badge turns red, trim the post before copying — usually cutting a phrase or dropping a hashtag is enough.
You get three posts per platform. Spread them out. Each one leads with a different angle, so staggering publication keeps your feed from feeling repetitive.
The tool writes copy only — it doesn't know your published URL. Paste the article link into each post before scheduling. On X, the URL counts toward your 280-character limit.
AI-generated hashtags are a reasonable starting point, but you know which tags your audience actually follows. Swap generic ones for the specific communities Defense One reaches.
LinkedIn posts get more engagement when they speak to a professional outcome — what does this mean for procurement, policy, or readiness? The AI tries, but you can sharpen it.
The tool avoids claims not in the article and won't use superlatives the story doesn't earn. That makes posts safer to publish quickly — but you can still add context, a reaction quote, or a sharper angle that the AI missed.
Most mobile email clients truncate at 40–45 characters. A subject fully visible without an ellipsis gets more opens. The badge turns green at 50 chars — tighter is almost always better.
Subscribers already trust you — don't tease. "Army cuts 3,000 civilian jobs" outperforms "Big changes coming to the Army." Names, numbers, and outcomes stop the eye.
Preview text is the gray snippet after the subject in most inboxes. Explicitly setting it to extend (not repeat) the subject is one of the highest-return, lowest-effort email improvements. Click the preview text in the results to copy it separately.
"Is the Pentagon ready?" underperforms for news alerts. Readers want to know what happened, not be invited to wonder. The tool avoids question subjects by design.
"$18B", "Hegseth", "F-35" are visually distinct and signal specificity. If the article has a compelling number or name, make sure it's in the subject — the tool will usually surface it, but check.
Subscribers already know they're reading Defense One. "Defense One: Pentagon freezes hiring" wastes 15 characters on information they already have. Start with the news.
"Search engines index headlines within minutes of publication. Getting the keyword right at launch matters more than any later correction." General principle, search optimization practice
| Element | Best Practice | Why It Matters |
|---|---|---|
| Length | 50–60 characters | Google Search typically shows ~60 chars; staying under prevents truncation with an ellipsis, which reduces click-through. |
| Keyword placement | As close to the start as natural | Search engines weight words earlier in the headline more heavily. Readers also scan-read left to right and need the hook fast. |
| Numbers | Use digits, not words ("5" not "five") | Numerals are visually distinct in a list of search results and signal specificity, which improves click-through rates. |
| Voice | Active preferred | Active constructions are shorter and more direct, which helps both SEO and reader comprehension. |
| Accuracy | Headline must reflect article content exactly | Clickbait headlines that don't deliver trigger high bounce rates, a negative ranking signal. This tool extracts explicit facts first and requires every headline to trace back to something in the article. |
| Freshness signals | Include year or timeframe for evergreen topics | "Best mortgage rates 2025" outperforms "Best mortgage rates" for date-sensitive searches. |
"Each platform is a different conversation. The same story needs a different opening line on LinkedIn than it does on X." General principle, social media practice
| Platform | Best Practice | What the Tool Does |
|---|---|---|
| 1–3 sentences; conversational; no hashtags; invite discussion | Writes in an accessible, engaged tone aimed at a general but news-literate audience. Focuses on why the story matters to readers' lives or interests. | |
| X | Under 280 chars; front-loaded hook; 1–2 hashtags | Writes punchy, compressed copy with the news value in the first few words. Includes relevant hashtags — check and swap these before posting. Badge turns red if a post needs trimming. |
| 2–4 sentences; professional tone; explain implications | Writes for defense professionals, policy analysts, and government audiences. Frames the story's significance for the industry or policy landscape rather than just summarizing the news. | |
| All platforms | Vary the angle across posts | The tool deliberately varies opening hooks, framing, and emphasis across the three posts per platform so you have options — and so your feed doesn't look repetitive if you post multiple times. |
| All platforms | Add your URL manually | The tool generates copy only. Paste each post, then add the published article link. On X, the link counts toward your 280-character limit. |
"The subject line is your headline. The preview text is your subhed. Most senders write neither with enough care." General principle, email marketing practice
| Element | Best Practice | What the Tool Does |
|---|---|---|
| Subject length | 40–50 characters (mobile sweet spot); 75 max | Targets under 50; badge is green at ≤50, neutral up to 75, red over 75. Tighter is almost always better. |
| Opening words | Lead with the news — name, number, or outcome | Avoids openers like "Defense One:", "BREAKING", or vague phrases. Instructs the AI to front-load specifics. |
| Question form | Avoid — underperforms for news alerts | Explicitly excluded from the prompt. The tool won't generate question-form subjects. |
| Preview text | 40–80 chars; extends subject without repeating it | Generated alongside each subject. Click the preview text in results to copy it separately from the subject line. |
| Voice | Active, past tense for completed events | Defaults to active constructions; avoids passive ("cuts were made") and nominalization ("a reduction of"). |
| Approach variety | Test different angles — open rates vary by framing | Generates 5 subjects with distinct approaches: straight news, key number/name, implication, urgency/consequence, curiosity/contrast. Use whichever fits the story best. |