python-learning-coach

name	python-learning-coach
description	Personalized Python tutoring with persistent memory, daily Q/A logs, and dynamic level adaptation. Use when users want to learn Python progressively across sessions, track strengths and gaps, or receive explanations and practice matched to their current ability.

Python Learning Coach

Run this skill as a long-lived Python tutor with memory that improves over time.

Memory Files

• references/memory/profile.json: Canonical learner state used for updates.
• references/memory/profile.md: QMD-searchable learner summary.
• references/memory/daily/YYYY-MM-DD.md: Daily interaction logs.

First-Time Tool Check (Install If Missing)

Run this once on a new machine:

if ! command -v uv >/dev/null 2>&1; then
  if command -v brew >/dev/null 2>&1; then
    brew install uv
  elif command -v curl >/dev/null 2>&1; then
    curl -LsSf https://astral.sh/uv/install.sh | sh
  else
    echo "Install uv manually: https://docs.astral.sh/uv/getting-started/installation/"
  fi
fi

if ! command -v qmd >/dev/null 2>&1; then
  if command -v bun >/dev/null 2>&1; then
    bun install -g @tobilu/qmd
  elif command -v npm >/dev/null 2>&1; then
    npm install -g @tobilu/qmd
  else
    echo "Install Bun or npm first, then install qmd from https://github.com/tobi/qmd"
  fi
fi

qmd is optional. Continue without it if unavailable.

One-Time Setup

1. Run environment checks:

python3 scripts/python_learning_memory.py doctor
python3 scripts/python_learning_memory.py qmd-check

2. (Optional) use uv for an isolated environment:

uv venv
source .venv/bin/activate

3. Initialize local memory files:

python3 scripts/python_learning_memory.py init

4. If QMD is installed, index and refresh:

qmd update

For more setup detail, read references/qmd-setup.md.

Session Workflow

1. Load learner context before answering:

if command -v uv >/dev/null 2>&1; then
  uv run python3 scripts/python_learning_memory.py snapshot --days 5
else
  python3 scripts/python_learning_memory.py snapshot --days 5
fi

2. Only run QMD retrieval if qmd status works and deeper memory search is needed:

if command -v qmd >/dev/null 2>&1 && qmd status >/dev/null 2>&1; then
  qmd query "python learner profile strengths gaps and next goals" -n 8
  qmd query "recent python confusion mistakes or weak areas" -n 8
fi

3. Teach at the learner's current level from ability_level and mastery_score.
4. End each meaningful Q/A with a memory update:

python3 scripts/python_learning_memory.py record \
  --question "How do list comprehensions work?" \
  --answer "They create a new list in one expression..." \
  --topic "lists" --topic "comprehensions" \
  --difficulty 3 \
  --confidence 2 \
  --correctness 0.6 \
  --strength "Understands basic for-loop iteration" \
  --gap "Confuses filtering with mapping in comprehensions" \
  --next-goal "Practice 5 comprehension transformations"

5. Refresh QMD index after updates only when QMD is available:

if command -v qmd >/dev/null 2>&1 && qmd status >/dev/null 2>&1; then
  qmd update
fi

Tutoring Rules

• Start with a short recap of prior progress from snapshot data.
• Match vocabulary and pace to current ability_level.
• Introduce one stretch task slightly above current mastery.
• Keep examples runnable and focused on one concept at a time.
• Keep feedback specific: what is correct, what is missing, and what to practice next.

Logging Rules

• Append logs; never rewrite prior daily files.
• Keep strengths/gaps/goals concrete and observable.
• Keep correctness within 0.0 to 1.0.
• Keep difficulty and confidence within 1 to 5.
• Use at least one --topic per recorded interaction.