Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    x-human: X character counter and AI-tell checker

    1

    A post that looks like 270 can be 296 and get cut off.

    Free

    Security scanned
    x-human: X character counter and AI-tell checker

    x-human: X character counter and AI-tell checker

    Example session with this skill installed

    Check this before I post it

    In today’s fast-paced world, I’m thrilled to share a game-changer — a seamless way to leverage AI for your X posts… It doesn’t just write. It elevates. Read more: https://github.com/Alok-Fusion/x-skills Thoughts?

    #AI #BuildInPublic #OpenSource #ClaudeCode #Productivity

    • Read your context and instructions
    • Compiled the x-human

    Checker output (real run of the bundled scripts/xcheck.py)

    X CHECK  ·  single post
    ----------------------------------------
     1  [255/280] OK   tags: #AI #BuildInPublic #OpenSource #ClaudeCode #Productivity
        fix   em dash x1  (--fix handles it)
        fix   curly apostrophe x3  (--fix handles it)
        fix   ellipsis character x1  (--fix handles it)
        flag  slop: leverage -> use, seamless -> smooth (or say what actually got easier), game-changer -> say what changed, thrilled to, thrilled -> glad / happy (or cut it)
        flag  bait: "Thoughts?"
        flag  hashtags 5, want 2-3 (more reads as spam)
        flag  link in the body: put it in a reply under the post
    ----------------------------------------
    VERDICT  FIX FIRST
    

    The link counts as 23, not its real length, so the post is 255 by X's count. Three mechanical tells can be fixed automatically. The slop words, the bait, the extra hashtags and the link need a rewrite.

    After the rewrite

    Wrote a checker that counts posts the way X does: every link is 23, every emoji is 2.
    
    It also catches em dashes, curly quotes and stock AI words before a post goes out.
    
    #BuildInPublic #AI
    
     1  [189/280] OK   tags: #BuildInPublic #AI
    VERDICT  READY
    

    Next step: post it, and put the repo link in a reply under the post instead of the body.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    What it does. Checks anything headed for X before you post it. X doesn't count characters the way len() does: every link counts as 23, every emoji as 2, and CJK as 2. A post that looks like 270 can be 296 and get cut off.

    How it works. The agent writes the draft to a file and runs the bundled scripts/xcheck.py. You get the real count, the headroom left, your hashtag count against your own rule, and flags for links in the body, em dashes, curly quotes, invisible characters, stock AI words (each with a plain alternative) and reflex bait like "Thoughts?". --fix cleans the mechanical tells. Wording flags get rewritten by the agent and checked again until the verdict is READY. Threads work too: posts separated by --- lines, with numbering and hashtag placement checked.

    Where it stops. These are local checks: X's public counting rules plus a word list. Nothing is uploaded. It can't tell you what an AI detector or X's ranking will do, and it never calls a post "undetectable".

    Part of x-skills, ten free MIT-licensed skills for X.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Social Media.

    Local checks only: X's public counting rules plus a list of common AI tells. It can't predict what an AI detector or X's ranking will do. The default limit is 280 characters (use --limit for other limits, like 160 for a bio). Needs Python 3.8+.

    How to use

    1. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    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

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed today

    • Free with an Agensi account

    Software
    Python 3.8+

    On Windows, if python3 is not found, run the checker with python instead.

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    File Scopes

    ~/.x-skills/voice.md
    ~/.x-skills/slop.json

    Runs the bundled scripts/xcheck.py on a draft saved to a local text file, and with --fix writes a cleaned copy. Reads ~/.x-skills/voice.md (your hashtag rule) and ~/.x-skills/slop.json (your own word list) if they exist. No network access.

    Listedtoday

    What's inside

    Frequently Asked Questions

    Models see tokens, not characters, and X doesn't count plain characters anyway: every link counts as 23, every emoji as 2, and CJK as 2. The bundled script follows X's published counting rules, so a post that looks like 270 but is really 296 gets caught before you post it.

    No. It checks text and prints a verdict (READY or FIX FIRST) with the real count. Posting is always up to you.

    No. x-human works on its own. The other nine skills (posts, threads, replies, planning, audits and more) are free on GitHub at github.com/Alok-Fusion/x-skills, and they run their drafts through this checker.

    Yes. Add your own words, or keep a flagged word you really use, in ~/.x-skills/slop.json. It uses the same format as the bundled assets/slop.json, plus an "allow" list.

    Only the mechanical tells: invisible characters, em and en dashes, curly quotes, the single-character ellipsis and non-breaking spaces. An emoji's own joiner is kept. It never changes your wording: slop words, bait, length and hashtag flags get rewritten by you or your agent, then checked again.

    New in Social Media

    Free