claudeers.
// Claude Plugins

english-exam-coach

Claude Code plugin for CEFR-aligned English exam prep: IELTS, TOEFL iBT (2026 format), CEFR B1-C2. Original content only; append-only progress tracking.

// Claude Plugins[ cli ][ api ][ desktop ][ web ][ claude ]#claude#agent-skills#cefr#claude-code#claude-plugin#english-learning#exam-preparation#ielts#plugins◷ MIT$open-sourceupdated about 1 month ago
Actively maintained
97/100
last commit about 1 month ago
last release about 1 month ago
releases 13
open issues 0
// star history

Install with your AI

Paste into Claude Code, Cursor, or any agent — it reads the repo and wires the tool into your project.

Install and set up english-exam-coach (claude-plugin project) into my current project.
Found on https://claudeers.com/english-exam-coach
Repo: https://github.com/OleksiiDotsenko/english-exam-coach
Homepage/docs: —
Detected install method: claude-plugin → /plugin install english-exam-coach@OleksiiDotsenko/english-exam-coach
Category: plugins. Platforms: cli, api, desktop, web.
Read the repo's README for exact setup and env vars, then install it and wire it into my project.

Claudeers Health Verdict:
active; community-verified: false. Confirm the source before running anything.
// or install directly (claude-plugin)
/plugin marketplace add OleksiiDotsenko/english-exam-coach
/plugin install english-exam-coach@OleksiiDotsenko/english-exam-coach
// or clone
git clone https://github.com/OleksiiDotsenko/english-exam-coach

// compatibility

Platformscli, api, desktop, web
Operating systems—
AI compatibilityclaude
LicenseMIT
Pricingopen-source
LanguagePython

Get your FREE $2.50 API credits to access TickAtlas financial data ↗

English Exam Coach — pixel banner

English Exam Coach turns Claude Code into an English exam tutor. It sets original practice in the exact format of your exam, grades it with feedback that tells you what to fix, and keeps a record on your own disk — so every session knows your trend, your weakest task type, and the one thing to drill next.

New in 3.0: it works for tutors too. Log a student's full test from any platform, review it against their earlier tests, see which mistakes are habits, and print practice sheets with answer keys.

[!IMPORTANT] Indicative, not official. All scores and level estimates are self-practice feedback — they are not official results and do not predict them. This project is not affiliated with, endorsed by, or connected to IELTS, ETS/TOEFL, Cambridge English, or any exam board; exam names are used nominatively to describe compatibility.

🔁 It remembers what you got wrong

Most tools grade you and forget. This one keeps a private ledger of your actual mistakes, schedules them for re-testing, and — the part nothing else does — writes a brand-new item on the same point when it comes back, so you can never answer from memory of the card.

You:    /daily-drill
Claude: Re-testing 3 points you have missed before — "past perfect after
        'by the time'" is due today; you have got it wrong twice.
        [three fresh items, none of them the ones you saw before]
You:    [your answers]
Claude: 2 right. "By the time" is promoted to next week; the article point
        comes back tomorrow. Attempt and errors logged.

Every score also decomposes: instead of "your writing is B2", the report tells you which criterion is holding it there, and says what the estimate rests on — a typed transcript and a real recording never count the same.

🧑‍🏫 For tutors: full tests, habits, paper

A single drill says how one task went. A run of full tests is the only honest trend line — and laying every mistake across that run is what separates a slip from a habit.

You:    /review-test Dana "Mock 6"      [score report + her writing attached]
Claude: Logged: R 5.5 · L 5.5 · W 5.5 · S 4.0 → overall 5.0, her record.
        4 mistakes in 4 points — 1 new, 3 seen before:
          missing article before a singular count noun   in 6 of 6 tests
          purpose questions answered with the topic      in 5 of 6 tests
          "adapted" confused with "adopted"              back after 3 clean tests
        Stayed away: third-person -s — 1 clean test, 3 to go.
You:    /worksheet Dana articles
Claude: sheet.html (with key) and sheet-learner.html written — open in a
        browser and print, or save as PDF.
  • coach log-test records a test's scores exactly as the test gave them.
  • coach tests shows the history: records, first half against second half, per-task detail — and refuses to call two tests a trend.
  • coach catalog is the error catalog: every mistake ranked, a point-by-test matrix, mistakes repeated word for word, mistakes carried over from the first language, and a habit state for every point. One rule throughout: a point is closed only after four clean tests in a row. Two clean tests is not closed.
  • coach render turns any report or worksheet into a page that prints cleanly — answer key on its own page, or left out for the learner's copy.
  • --learner <name> keeps each student in a folder of their own.

🧡 What you can do

  • 📝 Writing — paste your essay, email or report; get a band range, a CEFR level, and rewrites built from your own sentences
  • 🗣️ Speaking — exam-format prompts under the real clock; feedback on a transcript, or on a recording if you have a local speech recogniser
  • 📖 Reading & Use of English — cloze, key word transformations, matching, True/False/Not Given… scored objectively, every answer explained
  • 🎧 Listening — original scripts with exam-style questions, spoken by your computer's own voices on macOS and Linux, read-once elsewhere
  • 🧠 Vocabulary — levelled word sets with spaced repetition
  • 🗓️ Study plan — tell it your exam date; get a week-by-week plan it checks you against
  • 📈 Progress — every attempt logged locally; reports show trends and pick your next drill
  • 🧾 Full-test review, error catalog, printable sheets — see above

🚀 Get started in 2 minutes

You need Claude Code and Python 3.8 or newer. Most Linux distros include it; on macOS you may be prompted to install the Command Line Tools the first time (xcode-select --install) or grab it from python.org; on Windows, install from python.org.

1. Open a terminal and start Claude Code:

claude

2. Inside Claude Code, run these two commands one at a time — paste the first, press Enter, wait for the ✔, then paste the second:

/plugin marketplace add OleksiiDotsenko/english-exam-coach
/plugin install english-exam-coach@english-exam-coach

Using the desktop or web app — or seeing "/plugin isn't available in this environment"? Skip step 1 and run both commands from a plain terminal instead, prefixed with claude:

claude plugin marketplace add OleksiiDotsenko/english-exam-coach
claude plugin install english-exam-coach@english-exam-coach

3. Start a new session and type:

/start-prep

That's it — it asks which exam you're preparing for, finds your level, and starts your first task. Prefer talking? Just say "help me prepare for TOEFL" or "here's my IELTS essay: …".

Say / runWhat happens
/start-prepGuided start: exam → level → section → practice
/assess-level toefl-ibt15-minute CEFR check, logged as your baseline
"Give me a C1 key word transformation drill"Original items, objective scoring, explanations, logged
"Here's my IELTS Task 2 essay: …"Criteria-based feedback, band range + CEFR level, prioritised rewrites, logged
/mock-exam toefl-ibt fullA timed mock — one section or the whole test
/daily-drill10–15 minutes on what your own log says is weakest
/session-report · /progressReports for the session / all time
/review-testLog and review a full test taken elsewhere
/error-catalogWhat keeps going wrong across tests
/worksheetA printable practice sheet with an answer key

📚 Supported exams

exam idExamCEFR anchorScale
ielts-academicIELTS AcademicB1–C2Bands 0–9
ielts-generalIELTS General TrainingB1–C2Bands 0–9
toefl-ibtTOEFL iBT (2026 format)A1–C2Bands 1–6, half steps
cefr-b1B1 PreliminaryB1Cambridge English Scale
cefr-b2B2 FirstB2Cambridge English Scale
cefr-c1C1 AdvancedC1Cambridge English Scale
cefr-c2C2 ProficiencyC2Cambridge English Scale

Format facts (task types, item counts, timings, word counts, scales) live in data/exam-formats/. The TOEFL iBT file was re-verified in October 2026 against the exam provider's published test specifications and its full-length practice test; the task shapes it describes are pinned by the test suite. Formats change — confirm details with the exam provider before test day.

TOEFL item shapes are built by rule. Complete the Words and Build a Sentence have a mechanical shape that is easy to get subtly wrong by hand — a blank too many, a frame that does not match its tiles. coach ctest and coach sentence cut a plain paragraph or sentence into the exact shape and mark the answers, so an item is always answerable.

📈 Your progress, on your disk

Plain files in a folder of your own (never inside the plugin):

  • attempts.jsonl — one line per scored attempt.
  • errors.jsonl — the specific points you got wrong.
  • tests.jsonl — the scores of full tests.
  • queue.json — when each point is due to come back.
  • profile.json — your target exam, score and date.
  • reports/ — session reports, the all-time overview, test reviews and the error catalog, as Markdown (and HTML when you render them). Always regenerable.
  • learners/<name>/ — the same layout again, one per student, in tutor mode.

The three logs are append-only: nothing is rewritten behind your back, and a correction is a new line (--amend, --void) rather than an edit. JSON state is written atomically, and a file that somehow gets corrupted is set aside rather than silently discarded. coach state validate checks the directory; coach state export packs it into one zip to move or back up.

Cross-exam trends are normalised to CEFR — an IELTS band and a TOEFL score are never merged into one number.

Where the files live (first match wins): the --base flag → EXAM_COACH_HOME → ~/english-exam-coach/ (created on first use).

[!TIP] Obsidian user? Point EXAM_COACH_HOME at a folder inside your vault and the reports' YAML front matter (type, date, session, tasks, minutes, exams) makes Dataview tables and trend views work with no extra code.

❓ FAQ

Does it cost anything extra?

No separate account, API key or subscription — it runs inside your existing Claude session.

Where does my data go?

Nowhere. The plugin makes zero network calls and sends no telemetry. Your practice log is a set of local files in a folder you choose.

Will it predict my real exam score?

No — and be suspicious of anything that claims to. Estimates are honest ranges tied to public CEFR descriptors: good enough to steer your practice, not an official measurement.

I'm a tutor. Where do my students' results go?

Into learners/<name>/ inside your progress folder — one folder per student, never mixed, never uploaded. Use coach state export --learner <name> to hand a student their own history.

Can it make a PDF?

It makes a self-contained HTML page laid out for print. Open it in any browser and print it, or choose "Save as PDF" in the print dialog. No PDF library is bundled, by design: the scripts are standard library only.

Can it play listening audio?

On macOS (say) and Linux (espeak) it speaks scripts with the system voices, a different voice per speaker. Elsewhere listening drills fall back to read-once scripts — or bring your own audio and it builds questions for that.

Can it listen to my speaking?

It never records you. If you give it a recording and you already have a local speech recogniser installed (whisper-cli or whisper), it transcribes the file on your machine and measures pace and pauses. Otherwise you paste a transcript. Pronunciation is never scored.

The commands don't show up after installing

Start a new Claude Code session (or restart the app) — plugins load at session start.

"SSH authentication failed" when adding the marketplace

Two common causes. (1) Both install lines were pasted at once — the second line becomes extra arguments of the first and mangles the repository address. Run the commands one at a time. (2) If a clean, single command still tries SSH, add the marketplace by its HTTPS URL instead: /plugin marketplace add https://github.com/OleksiiDotsenko/english-exam-coach. No SSH key is ever required — the repository is public.

"/plugin isn't available in this environment"

The /plugin dialog only exists in the terminal version of Claude Code. In the desktop or web app, run the equivalent commands from any terminal: claude plugin marketplace add OleksiiDotsenko/english-exam-coach, then claude plugin install english-exam-coach@english-exam-coach.

How do I update or uninstall?

Update: /plugin marketplace update english-exam-coach, then /plugin update english-exam-coach@english-exam-coach (the full name@marketplace form is required here). Uninstall: /plugin uninstall english-exam-coach. Your progress files are yours and are never touched by either.

Upgrading from 2.x

Your progress folder is read as it is — the logs are additive-only by contract. What changed is the packaging: the eight skills are now one, english-exam-coach, and the commands are unchanged. Nothing to migrate.

🔒 Security & privacy

  • No network access. No network calls, no telemetry, nothing downloaded. Everything runs locally in your Claude session.
  • No MCP servers, no hooks, no background processes.
  • One bookkeeping tool, provably inert. coach.py and everything it can import use the Python standard library only, start no other program and open no connection. That is not a promise but a test: the suite walks the import graph and fails on any module that could.
  • It writes in one place. Your progress directory, plus a page or an archive you ask for. The logs are opened in append mode only.
  • Harmless wherever it is pointed. A learner's essay is untrusted text. coach.py reads only working files (.txt, .md, .json, .jsonl, links followed), writes pages only to .html and archives only to .zip, and replaces an existing file only if it wrote it. Export and import move the files of the progress layout and nothing else. No tool is pre-approved — you decide what runs without asking.
  • Three optional helpers start programs already on your machine, and only when a task needs them: speak.py and timed_speak.py (the system speech synthesiser and an audio player), and transcribe.py (a speech recogniser you installed, on a recording you provide). Audio is rendered to a temporary folder and deleted after the drill. Nothing records you.
  • Nothing a report contains can run. The print renderer escapes all HTML, loads no scripts, fonts or images, and links only to ordinary web and mail addresses.

🧩 What's inside

  • One self-contained skill, english-exam-coach: a short router (SKILL.md), a reference file per area loaded on demand, reference data, and the scripts — all in one folder, so it works wherever a skill's own folder is all there is.
  • 9 commands: /start-prep, /assess-level, /daily-drill, /mock-exam, /session-report, /progress, /review-test, /error-catalog, /worksheet.
  • Data: exam format facts, paraphrased public CEFR descriptors, calibration anchors, a closed error taxonomy, and original seed items used as format references.
  • Scripts: coach.py and the modules behind its 15 commands, plus the three optional helpers. Python 3 standard library only.

⚖️ Content and IP policy

  • No official exam content. No past papers, official item banks, or official mark schemes anywhere in this repository. All practice items are original, generated to match public format facts (task types, counts, timings, scales — facts are not copyrightable). Task instructions are written in our own words.
  • Public descriptors only. Evaluation anchors are condensed and adapted from the public CEFR framework (© Council of Europe, freely published); criterion names are used nominatively. Some short phrases stay close to the concise source wording.
  • Brand-neutral naming. Files and identifiers are named by CEFR level where possible; trademarks appear only nominatively, with no claim of affiliation or endorsement.

If you believe any content in this repository crosses these lines, please open an issue — it will be treated as a bug.

🛠️ Development

python3 -m unittest discover -s tests      # from the repo root
claude plugin validate . --strict
claude plugin validate ./plugins/english-exam-coach --strict
python3 tools/make_banner.py               # regenerate assets/banner.svg
python3 tools/make_icon.py                 # regenerate the listing icon

Design notes: one skill routes to reference files; exam differences are data, not code; the logs are append-only and every report is derived from them. DOGFOOD.md lists what the tests cannot prove. See CHANGELOG.md for release history.

License

MIT © 2026 Oleksii Dotsenko

// faq

What is english-exam-coach?

Claude Code plugin for CEFR-aligned English exam prep: IELTS, TOEFL iBT (2026 format), CEFR B1-C2. Original content only; append-only progress tracking.. It is open-source on GitHub.

Is english-exam-coach free to use?

english-exam-coach is open-source under the MIT license, so it is free to use.

What category does english-exam-coach belong to?

english-exam-coach is listed under plugins in the Claudeers registry of Claude-compatible tools.

24 views
★ 11 stars
unclaimed
updated about 1 month ago

// embed badge

english-exam-coach on Claudeers
[![Claudeers](https://claudeers.com/api/badge/english-exam-coach.svg)](https://claudeers.com/english-exam-coach)

// retro hit counter

english-exam-coach hit counter
[![Hits](https://claudeers.com/api/counter/english-exam-coach.svg)](https://claudeers.com/english-exam-coach)

// reviews

// guestbook

0/500

// related in Claude Plugins

🔓

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

// pluginsmultica-ai/★ 230,631[ claude ]
🔓

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explainin…

// pluginsanthropics/⟨Python⟩★ 156,472[ claude ]
🔓

"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/

// pluginsHKUDS/⟨Python⟩★ 58,467◷ Apache-2.0[ claude ]
🔓

financial-services — a Claude ecosystem project on GitHub.

// pluginsanthropics/⟨Python⟩★ 41,409◷ Apache-2.0[ claude ]
→ see how english-exam-coach connects across the ecosystem