Grok Trend Angler

What it does

You give it your topic, your product, what you know first-hand and, if you like, what you want from posting. It reads how lively the conversation on X is and names what it based that on. Then it opens with its call: the post to publish now, written in full and ready to paste, and whether to post it on its own, as a reply or as a quote post. One or two alternatives on other angles follow. The whole answer is usually 150-300 words.

What it adds to a plain "give me post ideas about this trend":

• A decision first. The answer starts with the one post to publish, and says when a fading conversation is better sat out.
• Angles chosen for this conversation and your goal, from a menu (a contrarian point you can defend, a useful number, a practical how-to, your own result or mistake, a precise answer to a question people keep asking), not one of each by rule.
• The trend check has to show its evidence. If the model can't see current posts, it says "No live view" instead of pretending.
• An "If challenged" line under every post, so you never post a claim you can't defend.
• Numbers only from you, a named source or visible arithmetic. Anything else becomes [N] with a note on what to look up.
• Your product where it solves the problem people are discussing, backed by your own numbers when you have them, as the tool rather than the headline.
• Posts that read like posts: first person and concrete, with a hook in the first line, kept under X's length limit with a margin, and counted.

Use it when

• Your niche is being discussed on X today and you want to add something better than a quote-post.
• You post about a product and want posts where the product is the tool, not the headline.
• You have ten minutes and want one post you can publish now, plus a backup.

Skip it when

• You want a blog post or newsletter; this is built for short posts.
• There is no live conversation and you want evergreen posts for the weeks ahead. A content plan fits better.
• The news is a tragedy or a crisis outside your work. Don't angle it.

The prompt

You are a social strategist helping me join today's conversation on X about my topic in a way that serves my goal, without hijacking the conversation and without posting anything I can't back up.

My goal for posting may be in my inputs below (for example sign-ups for my product, replies from people in my field, or being known for this topic). If I did not state one, assume I want the people in this conversation who have the problem my product solves to notice it, without the post reading like an ad, and say so in the Assumed line. Do not ask me questions first.

Before writing, read the conversation:
- Is it active, steady or fading right now? Base that on 2-3 posts, accounts or news items you can name. If I pasted posts, they are your view of the conversation. If I pasted none and you cannot see current posts in this session, say "No live view" and write evergreen posts.
- What do people keep complaining about, asking or getting wrong?
- Which angles fit this conversation and my goal? Pick the strongest from this menu; no type is required:
  - Contrarian: a point most of the conversation misses, in one sentence I could defend.
  - Useful number: a figure people would save.
  - How-to: a practical step someone can use today.
  - My result: something I did, measured or got wrong.
  - Answer: a precise answer to a question people keep asking.

Then answer in this order:

1. POST FIRST: one line with your call (post now, post later today, or sit this one out, which is the right call when the conversation is fading or has turned to a tragedy) and why, tied to the conversation and my goal. Say whether to post it on its own, as a reply to a post you named, or as a quote post. Then the recommended post, written in full and ready to paste. If you advise sitting it out, give the post I could use later instead.
2. TREND CHECK: at most 2 lines: active, steady or fading, with the posts, accounts or news items you based it on, or "No live view".
3. ALSO GOOD: 1 or 2 alternative posts on other angles, only if they are strong. Start each with its angle's name from the menu.
4. Closing lines, only where they apply:
   - "Don't repeat: <a wrong claim going around in the conversation, and why it is wrong>"
   - "Later today (optional): <one reply to add under the recommended post that says something the post does not>"
   - "Assumed: <your assumptions in one line>"

For every post:
- Write it the way a person posts on X: first person, concrete, plain words, one idea. No analysis framing ("This highlights", "The key takeaway"), no thread numbering.
- The first line is the hook. It has to work on its own in the timeline, in about 90 characters or fewer.
- Length: at most 280 characters unless I give a different limit. Aim 10-15% under the limit, then count and put the count in parentheses after the post.
- Numbers come only from MY KNOWLEDGE, a source you name, or simple arithmetic from stated inputs (show the sum). Otherwise write [N] and say which figure I should look up.
- When my goal involves the product, connect it to the problem the conversation is about, and put my own numbers in the recommended post when they back it up. Name the product once, as the tool, not the headline. Leave it out of any post where it does not fit.
- On the next line write "If challenged:" and the fact, source or arithmetic the post rests on, or "UNVERIFIED, check before posting".
- Do not build a post on correcting someone else's mistake unless the conversation is about that mistake; use the "Don't repeat" line instead.

Rules: no hashtags, no engagement bait ("RT if", "Agree?"), nothing about tragedies, deaths or ongoing crises unless my topic is directly about them. Keep the whole answer to about 150-300 words: the posts are the deliverable, and every other line stays short.

MY TOPIC (a specific debate, policy or release, not a whole industry): [TOPIC]
MY PRODUCT (name and what it does, in one line; add your goal for these posts if you have one): [PRODUCT]
MY KNOWLEDGE (numbers, results or experience I can stand behind, or "none"): [KNOWLEDGE]
POSTS I'VE SEEN (optional: paste 5-15 recent posts, or "none"): [POSTS]

How to use

1. Make [TOPIC] specific: "four-day work week pilots", not "work".
2. Put anything you know first-hand in [KNOWLEDGE] (your own numbers, client results, a mistake you made). That is where the useful number and the defensible take come from.
3. Add your goal to the [PRODUCT] line, for example "goal: trial sign-ups from team leads". Without one, the prompt assumes you want the product noticed without an ad tone, and says so.
4. If your model can't see X, or you want the posts tied to particular posts, copy 5-15 recent posts on the topic into [POSTS].
5. Paste the prompt as your first message (see How to run).
6. Read POST FIRST and TREND CHECK. If TREND CHECK names posts, open two of them before you rely on them. If it says "No live view", the posts are evergreen drafts.
7. Fix every [N] and every UNVERIFIED line before posting: find the source, or drop the claim.
8. Publish the recommended post the way POST FIRST says (on its own, as a reply or as a quote). Keep the "Later today" reply for later in the day.

How to run (Oct 2026)

• Grok: paste it as the first message on grok.com, in the Grok apps or in X. Don't assume it can see live posts: check that TREND CHECK names real posts you can open. Through the xAI API, Grok 4.7 defaults to high reasoning_effort; medium is enough here.
• Claude, ChatGPT, Gemini: paste it as the first message and fill [POSTS] with recent posts you copied from X, so the trend check has something real to read. Without them you get "No live view" and evergreen posts, which is the honest result.
• ChatGPT: Instant or Medium on the Thinking slider is enough. Claude: Claude Sonnet 5.5 is quick for this short job. Gemini: in the Gemini app, you can save the instructions part as a skill (type / and its name) and send the four input lines each time.
• Any model: check post lengths with a character counter; model counts are estimates.

Example

Input: topic: four-day work week pilots · product: Notewell, a meeting-notes app that turns calls into action lists; goal: trial sign-ups from team leads · knowledge: our 8-person team cut recurring meetings from 9 to 4 a week last quarter · posts: 12 pasted posts from founders and managers

Output (excerpt)

POST FIRST: Post now, as a reply to the manager in your pasted posts who doubts a team can cover a four-day week. Your own meeting numbers answer that doubt, and team leads are the people reading that thread.

Four-day weeks don't fail because of the day off. They fail because the meetings stay.
Our 8-person team cut recurring meetings from 9 a week to 4 last quarter. Cut meetings first, then the day. I build Notewell, which turns calls into action lists. (249 characters)
If challenged: your own meeting count from last quarter.

TREND CHECK: Active, based on the posts you pasted: two pilot results, the manager's doubts about coverage, a founder's "never again" post.

ALSO GOOD
Useful number: One weekly 60-minute meeting with 8 people is 8 person-hours. On a 32-hour week, that is a quarter of one person's week. Every week. (132 characters)
If challenged: arithmetic from the inputs shown (60 min × 8 = 8 hours; 8 ÷ 32 = 25%); no survey figure claimed.

Later today (optional): reply with the first recurring meeting your team cut and what took its place.

Illustrative example written by Meevra to show the target format; your model's wording will differ.

Troubleshooting

• TREND CHECK names posts you can't find. Treat the result as "No live view": paste real posts into [POSTS] and run it again.
• The useful number has no source. Reply "Where does this figure come from? Name the source or replace it with [N]." Arithmetic from stated inputs is safer than survey figures.
• The product shows up in every post. Reply "Mention the product only where it solves the problem this conversation is about; take it out of the others."
• A post reads like analysis. Reply "Rewrite it as a post: first person, one concrete point, no 'this highlights'."
• A post is over the limit. Count it yourself, then reply "Cut this post to about 240 characters and keep the number."
• The posts feel generic. Narrow [TOPIC] to one event, policy or release, and add something you know first-hand to [KNOWLEDGE].

© 2026 Meevra · meevra.com · Licensed to the purchaser for personal and internal business use.