Run this task with `hai` (https://github.com/braincore/hai-cli).
Install `hai` from your terminal: curl -LsSf https://hai.superego.ai/hai-installer.sh | sh
Run this task from your terminal: hai task ken/flashcard-review

---

name = "ken/flashcard-review"

version = "1.1.0"

description = "Reviews flashcards."

steps = [
    "/asset-temp flashcard/deck 50",
    """/prep This task helps the user review "flashcards" (questions & answers)
that they have previously generated.

The cards are stored in an asset (`flashcard/deck`). Cards from the deck have
already been downloaded locally to temporary files. Each temporary file
contains one flashcard.

Each flashcard is of the following format:

```
{{Question}}

> {{Answer}}
```

When providing only the question, the answer lines (>) should be removed.
""",
"!fn-py(cache=true) Write a function that picks a random number [0, N).",
"!fn-py(cache=true) Write a function that that takes a file path to a flashcard with the above format. It should read the file and return ONLY the question.",
"!fn-py(cache=true) Write a function that that takes a file path to a flashcard with the above format. It should read the file and return ONLY the answer with > removed.",
"""/prep When asked to review a card using the !hai tool, you should return the
following subcommands:

    1. /f0 pick a random card from the deck
    2. /f1 <quoted path of card indexed by the random number>
       - Prints the question
    3. /ask-human Answer?
    4. /f2 <quoted path of card indexed by the random number>
       - Prints the answer
    5. /prompt Evaluate the answer

Afterwards, the user may ask you to review more cards. In that case, propose the same
subcommands
""",
"!hai",
"!hai review a card",
]