More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Plain English Writing (Simplified Technical English)

    1

    Your agent writes checklists, guides and emails in plain English, then checks every sentence with a script.

    $9

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Plain English Writing (Simplified Technical English)

    Plain English Writing (Simplified Technical English)

    Example session with this skill installed

    Rewrite these setup steps as a plain English checklist and run the checker on it.

    • Read your context and instructions
    • Compiled the plain english writing

    Here is the checklist. The 7 original steps became 16, with one action each and every fact kept. I moved the label loading ahead of the test print, since the labels have to be in first.

    1. Download the driver from the support page.
    2. Save the file in a folder you can find again.
    3. When the download finishes, double-click the installer.
    4. Do not connect the printer yet. Connecting it too early is the most common reason the setup fails.
    5. On the welcome screen, click Next.
    6. Accept the license agreement.
    7. If the installer asks for a setup type, choose Typical.
    8. When the installer tells you to connect the printer, connect the USB cable to the computer and the printer.
    9. Turn the printer on. The installer finds it automatically.
    10. When the installer says it is complete, click Finish.
    11. If the installer asks for a restart, restart the computer.
    12. In System Settings, open Printers & Scanners.
    13. Check that the LP-200 is in the list.
    14. If the LP-200 shows as offline, reconnect the USB cable.
    15. Load the labels with the shiny side facing up.
    16. Print a test label from the printer's Options page to confirm that it works.

    Checked as steps (--surface steps):

    OriginalRewrite
    FAIL lines200
    NOTE lines10
    Steps passing0 of 716 of 16
    Words193173

    The 20 FAILs in the original were 6 LONG, 6 PASSIVE, 4 MULTI, 3 CONDITION and 1 SEMICOLON. The NOTE was the trailing ", making sure..." clause in the last step.

    Nothing is left in the report, but two things are worth knowing. My first rewrite also came back clean while step 4 said the setup "usually fails", which is not what the original claimed. A read-through caught it, so the wording now matches. The checker also missed one problem in the original: step 1 bundled "then double-click the installer" without a MULTI flag.

    Next steps

    • Confirm Options is the name the printer's own page shows.
    • Check that "reconnect the USB cable" in step 14 reads clearly to your users.

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

    About this skill

    Most AI agents write long sentences and steps that bundle three actions into one line. Readers miss the step that mattered and come back with questions. This skill gives your agent clear plain-English rules and a checker that catches slips before anyone reads the draft.

    The rules come from ASD-STE100 Simplified Technical English, the controlled language of aircraft maintenance manuals. You get them at three strengths:

    • Strict for checklists and procedures: one command per sentence, 20 words at most, the condition first, active voice, the same name for each thing every time.
    • Compact for guides and install instructions: short commands, with each fix grouped under what the reader sees on screen.
    • Light touch for emails, chat replies and reports: sentences of 25 words or fewer, short paragraphs, the actor named, no semicolons.

    The checker is a single Python 3 script with no dependencies and no network access. Your agent runs it on a draft and gets a report of each rule break. Each line quotes the sentence at fault, so the agent can fix it and check again. It also flags file paths and shell commands in replies meant for non-technical readers. Any number missing from a facts file you supply gets flagged for a check. It ships with 71 tests.

    Your own banned-word list is an editable block at the top of the script, or a plain text file passed on the command line. The checker flags em and en dashes by default. To allow them, delete the two entries in BANNED_CHARACTERS at the top of the script. Contractions are allowed on purpose, which departs from strict STE.

    You need Python 3.9 or newer. It is English only and does not judge whether your facts are right. Not affiliated with or endorsed by ASD.

    English only. The checker works from patterns and word lists, so it cannot tell whether a fact is right, and it can miss a second instruction bundled into a step. It is based on ASD-STE100 but is not certified, and it allows contractions on purpose.

    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 any payment.

    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

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 13 days ago

    Requires

    Python 3.9 or newer
    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed13 days ago

    What's inside

    Frequently Asked Questions

    Python 3.9 or newer. The checker is one script that uses only the standard library, so there is nothing to install. It was tested on Python 3.9.6 and 3.14.0. If `python3` is not found on your machine, try `python`.

    No. The checker makes no network calls and reads no environment variables. It reads the draft from a file or from standard input and prints a report, as text or as JSON with `--json`. Your agent does the rewriting in your own session.

    Agents tend to write long sentences and steps that bundle several actions together. Here a script counts. It flags any step over 20 words, a condition placed after the command, passive voice, semicolons and a second instruction, and it quotes the sentence at fault. The agent fixes the draft and runs the check again, for at most 3 rounds.

    It works from regular expressions and word lists, so it cannot tell whether a sentence is clear or a fact is right. In our demo run a rewrite passed clean even though one step had changed a fact; a read-through caught it. It also missed one bundled instruction in the original. It reads English only.

    It ships with 71 tests, and all 71 passed in our demo run. In that run, seven messy printer setup steps scored 20 FAIL lines and 0 of 7 passing steps. The 16-step rewrite scored 0 FAIL and 16 of 16.

    It is based on ASD-STE100 but is not certified, and ASD does not endorse it. Contractions are allowed on purpose. You can edit the banned phrases, word limits and product names in a CONFIG block at the top of the script, or pass your own banned list as a text file.

    New in Writing & Content