learn-stats

# Learn Stats Reads `workspace/learning/facts/` and `workspace/learning/.state/review-log.jsonl` to produce a markdown stats report. ## Trigger User wants a snapshot of their learning loop health. ## Workflow ### Step 1 — Read facts 1. Read all `.md` files in `workspace/learning/facts/` 2. Parse frontmatter of each file 3. Get today's date (YYYY-MM-DD) 4. Compute: - `total` = count of all fact files - `overdue` = count where `next_review <= today` - `active_decks` = count of unique `deck` values across all facts - `added_this_week` = count where `created >= today - 7 days` If `workspace/learning/facts/` does not exist or is empty: > "Nenhum fato encontrado. Use /learn-capture para adicionar fatos primeiro." > Stop here. ### Step 2 — Read review log 1. Read `workspace/learning/.state/review-log.jsonl` line by line (each line is a JSON object) 2. If the file does not exist or is empty: `total_reviews = 0`, `retention_rate = "N/A"` 3. If it exists: - `total_reviews` = count of all log entries - `good_easy_count` = count where `grade == "good"` OR `grade == "easy"` - `retention_rate` = `round(good_easy_count / total_reviews * 100)%` ### Step 3 — Output report Print in markdown, in **pt-BR**: ```markdown ## 📊 Learning Loop — Stats | Métrica | Valor | |---------------------|--------------| | Total de fatos | {total} | | Vencidos hoje | {overdue} | | Taxa de retenção | {retention_rate} ({good_easy_count}/{total_reviews} revisões) | | Decks ativos | {active_decks} | | Adicionados (7 dias)| {added_this_week} | **Data:** {today YYYY-MM-DD} ``` If `overdue > 0`: > 💡 Use `/learn-review` para revisar os fatos vencidos. If `total < 10`: > 💡 Use `/learn-capture` para adicionar mais fatos (meta v0: 20 fatos em 2 semanas). ### Step 4 (optional) — List overdue facts If `overdue > 0` and `overdue <= 10`, list them: ``` ### Fatos vencidos - `{filename}` (deck: {deck}, venceu: {next_review}) > {Retrieval Q — first 80 chars} ``` ## Constraints - **Read-only.** Do NOT modify any file. - Do NOT write to review-log.jsonl. - If log is missing (user hasn't reviewed yet), show `N/A` for retention — do not error. - Round retention_rate to nearest integer (no decimal places).