agentleFS
liknenSliknenS/language-learning-skills/skills/exam-prep/SKILL.md

exam-prep

Skill0 starsChanged 25 days agoMITView on GitHub

Exam prep for a language exam (residence-permit and citizenship exams such as Lithuanian A2 and the Lithuanian Constitution test, the Polish B1 certificate, Czech A2 for permanent residence, or IELTS-style speaking) - format, timed mock tasks, examiner-style questions, feedback against the task. Use this skill whenever a language learner mentions an upcoming language exam, asks for a mock exam, wants to practise an exam part or asks what the examiner will ask. Also for "подготовка к экзамену", "підготовка до іспиту", "egzamin z polskiego".

9 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Full mock closer to the date: All parts in order, with the real timing, then a short report: what would pass, what to fix in the remaining …

    line 24
  2. Very specific:

    Start with today() and profile() if not done this session.

    line 29
  3. Distinctive:

    Run the drills in the chat as above.

    line 43

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Exam prep

Train the format as much as the language: timing, task types and what the examiner listens for.

Steps

  1. Pin down the exam. Which exam, level, part (oral, written, test), date. Check the current format on the official site of the exam body before drilling: formats change. For example: Lithuanian state language and Constitution exams - Nacionalinė švietimo agentūra (NŠA); Polish certificate exams - the State Commission (certyfikatpolski.pl); Czech A2 for permanent residence - NPI ČR (cestina-pro-cizince.cz).
  2. Map the parts. For each part: what the task looks like, preparation time, answer time, how it is scored. Tell the learner what you know for sure and what you assume.
  3. Drill one part at a time, timed. Give the task, say "you have N seconds", let the learner answer (spoken if the host allows, typed otherwise), then give feedback against the task: did they cover every point, length, range of vocabulary, the 2-3 mistakes that cost points.
  4. Build the exam kit. Topic vocabulary (see vocab-coach), ready phrases for gaining time and asking to repeat, a few model answers the learner adapts to their own life.
  5. Full mock closer to the date. All parts in order, with the real timing, then a short report: what would pass, what to fix in the remaining days. A mock is a forecast, not a result.

If Talkfirst MCP is connected

  • Start with today() and profile() if not done this session. Record the goal with profile_update when the learner asks: purpose {kind: "life", life: ["exam"]}, target_level, deadline; then docs("plan") and plan_set for the weeks left (see study-planner).
  • Mock exams run in the Talkfirst app, not in the chat: Lithuanian A2 (oral and written parts) and the Lithuanian Constitution test, Polish B1 oral, Czech A2 oral (beta). The app voices the tasks, times every part and, for the oral and written parts, gives an AI forecast of the score (a forecast, not the exam's decision; pronunciation is not graded). Send the learner to the Exam screen of the app (web https://app.talkfirst.io/app or the Telegram Mini App, the same account). On the free plan a full mock exam is available once in 7 days; parts can be trained within the free speaking minutes.
  • In the chat, prepare for it with the tools: topics() and vocab_add for the exam's topics, quiz_start on them, practice_phrase for the phrases the learner will say, and dialogue_start with a brief in the examiner's role (on the free plan only ready scenes: dialogue_start(scene=...)).

Without it

Run the drills in the chat as above. Spoken answers can be typed or dictated; be clear that you grade content and language, not pronunciation, unless a scoring tool is available.

Timed mock exams with an examiner voice and a forecast, plus the vocabulary and plan on one account: connect Talkfirst (free plan, MCP connector for Claude and ChatGPT).

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.