algernon-review

You run the interactive flashcard review session with FSRS-4.5 spaced repetition.
You handle flashcards (binary reveal), dissertative cards (AI-graded), and
argumentative cards (AI-graded). At the end, you check promotion eligibility and
save the session.

Constants

CODEBLOCK0

FSRS-4.5 Parameters

- DECAY = -0.5, FACTOR = 0.2346
Stability (S) = days to reach 90% retention
Grades: 1 = Again, 3 = Good

Step 1 — Fetch Due Cards

CODEBLOCK1

(Include AND m.slug = 'SLUG' only if a specific slug was provided.)

If no cards due: "No cards due for review. Great job staying on top of it." and stop.

Display: "Starting review: N cards due."

Step 2 — Review Loop

Flashcards (type = 'flashcard')

1. Show front. AskUserQuestion options: ["Show answer"]
Show back. AskUserQuestion options: ["Again", "Good"]
Run FSRS update (see Step 3).

Dissertative and Argumentative Cards

1. Show front. AskUserQuestion options: ["Ready to answer"]
AskUserQuestion: "Type your answer:" (free text)
Evaluate the response against the reference answer (card back):

- Dissertative: check accuracy of key points, completeness - Argumentative: check that both sides are represented, trade-offs identified - Output: brief feedback + suggested grade (1 or 3) + optional MISCONCEPTION note

4. Show evaluator feedback + reference answer. AskUserQuestion options: ["Again", "Good"]

(Use the user's button choice, not the AI suggestion.)

5. Run FSRS update using the user's chosen grade.
If a MISCONCEPTION was detected, create a correction card:

CODEBLOCK2

Step 3 — FSRS Scheduling

For each graded card, compute new values and update card_state.

Read current state:

CODEBLOCK3

Compute elapsed days (if last_review is not NULL):

CODEBLOCK4

State transitions:

State	Grade	New stability	New difficulty	New state	Interval
new	Good	0.4	0.3	review	1 day
new

For review+Good, compute retrievability first:
CODEBLOCK5

Update:

CODEBLOCK6

Step 4 — Promotion Check (after all cards)

For each card reviewed with grade = Good where reps >= 5:
CODEBLOCK7

If reps >= 5 and tags contain [N1], check deck retention over last 7 days:
CODEBLOCK8

If retention >= 0.9:

- Generate a deeper N2 version of the card (N2: differentiator + when to use + main trade-off).
Insert as new card with tag [N2], due today.
Apply same logic for [N2] cards: promote to N3 (full technical depth, production

nuances, edge cases).

Step 5 — Session Summary

CODEBLOCK9

Append to today's conversation log:
CODEBLOCK10

algernon-review

你使用FSRS-4.5间隔重复算法运行交互式闪卡复习会话。
你处理闪卡（二进制揭示）、论述卡（AI评分）和
论证卡（AI评分）。最后，你检查升级资格并
保存会话。

常量

bash
ALGERNONHOME=${ALGERNONHOME:-$HOME/.openalgernon}
DB=${ALGERNON_HOME}/data/study.db

FSRS-4.5 参数

- DECAY = -0.5, FACTOR = 0.2346
稳定性（S）= 达到90%保留率所需天数
评分：1 = 再来一次，3 = 良好

步骤 1 — 获取到期卡片

bash
sqlite3 $DB \
SELECT c.id, c.type, c.front, c.back, c.tags, c.sourcetitle, c.deckid,
cs.stability, cs.reps, cs.state
FROM cards c
JOIN cardstate cs ON cs.cardid = c.id
JOIN decks d ON d.id = c.deck_id
JOIN materials m ON m.id = d.material_id
WHERE cs.due_date <= date(now)
[AND m.slug = SLUG]
ORDER BY cs.due_date ASC
LIMIT 50;

（仅当提供了特定slug时，包含AND m.slug = SLUG）

如果没有到期卡片：没有到期需要复习的卡片。做得很好，保持领先。并停止。

显示：开始复习：N张卡片到期。

步骤 2 — 复习循环

闪卡（type = flashcard）

1. 显示正面。AskUserQuestion选项：[显示答案]
显示背面。AskUserQuestion选项：[再来一次, 良好]
运行FSRS更新（见步骤3）。

论述卡和论证卡

1. 显示正面。AskUserQuestion选项：[准备回答]
AskUserQuestion：输入你的答案：（自由文本）
对照参考答案（卡片背面）评估回答：

- 论述卡：检查关键点的准确性、完整性 - 论证卡：检查是否呈现了双方观点，是否识别了权衡 - 输出：简短反馈 + 建议评分（1或3）+ 可选的误解备注

4. 显示评估反馈 + 参考答案。AskUserQuestion选项：[再来一次, 良好]

（使用用户选择的按钮，而非AI建议。）

5. 使用用户选择的评分运行FSRS更新。
如果检测到误解，创建纠正卡片：

bash sqlite3 $DB \ INSERT INTO cards (deck_id, type, front, back, tags) VALUES (DECK_ID, flashcard, 纠正：误解问题, 正确解释, [\[correction]\,\[N1]\]); INSERT INTO cardstate (cardid, due_date) VALUES (lastinsertrowid(), date(now));

步骤 3 — FSRS 调度

对于每张已评分的卡片，计算新值并更新card_state。

读取当前状态：

bash sqlite3 $DB \ SELECT stability, difficulty, reps, lapses, state, last_review FROM cardstate WHERE cardid = CARD_ID;

计算经过天数（如果last_review不为NULL）：

bash sqlite3 $DB \ SELECT ROUND(julianday(now) - julianday(LAST_REVIEW), 2) AS elapsed;

状态转换：

状态	评分	新稳定性	新难度	新状态	间隔
新	良好	0.4	0.3	复习	1天
新

再来 | 0.1 | 0.4 | 学习中 | 1天 | | 学习中 | 良好 | 稳定性 * 1.5 | MAX(0.1, 难度 - 0.05) | 复习 | MAX(1, ROUND(S)) | | 学习中 | 再来 | 稳定性（不变） | MIN(1.0, 难度 + 0.1) | 学习中 | 1天 | | 重新学习 | 良好 | 稳定性 * 1.5 | MAX(0.1, 难度 - 0.05) | 复习 | MAX(1, ROUND(S)) | | 重新学习 | 再来 | 稳定性（不变） | MIN(1.0, 难度 + 0.1) | 重新学习 | 1天 | | 复习 | 良好 | S EXP(0.9(1-R)) | MAX(0.1, 难度 - 0.05) | 复习 | MAX(1, ROUND(S)) | | 复习 | 再来 | MAX(0.1, 稳定性 * 0.2) | MIN(1.0, 难度 + 0.1) | 重新学习 | 1天，失误+1 |

对于复习+良好，先计算可检索性：
bash
sqlite3 $DB \
SELECT EXP(LN(0.9) * ELAPSED / STABILITY) AS R;

更新：

bash sqlite3 $DB \ UPDATE card_state SET stability = NEW_S, difficulty = NEW_D, due_date = date(now, + || INTERVAL || days), last_review = datetime(now), reps = reps + 1, lapses = NEW_LAPSES, state = NEW_STATE WHERE cardid = CARDID; INSERT INTO reviews (cardid, grade, scheduleddays, elapsed_days) VALUES (CARD_ID, GRADE, INTERVAL, ELAPSED);

步骤 4 — 升级检查（所有卡片之后）

对于每张评分良好且reps >= 5的卡片：
bash
sqlite3 $DB \
SELECT c.id, c.tags, c.deck_id, cs.reps
FROM cards c JOIN cardstate cs ON cs.cardid = c.id
WHERE c.id = CARD_ID AND cs.reps >= 5;

如果reps >= 5且标签包含[N1]，检查过去7天的牌组保留率：
bash
sqlite3 $DB \
SELECT CAST(SUM(CASE WHEN grade=3 THEN 1 ELSE 0 END) AS REAL) / COUNT(id) AS retention
FROM reviews r JOIN cards c ON c.id = r.card_id
WHERE c.deckid = DECKID AND r.reviewed_at >= datetime(now, -7 days);

如果保留率 >= 0.9：

- 生成卡片的更深层次N2版本（N2：区分因素 + 使用时机 + 主要权衡）。
作为新卡片插入，标签为[N2]，今天到期。
对[N2]卡片应用相同逻辑：升级到N3（完整技术深度、生产细节、边缘情况）。

步骤 5 — 会话摘要

会话完成。
复习卡片数：N
再来一次：X | 良好：Y
本次会话保留率：Z%
下次复习：[来自card_state的最早到期日期]

追加到今天的对话日志：
bash
echo [HH:MM] review session | Cards: N | Retention: Z% | Promotions: P \
>> ${ALGERNON_HOME}/memory/conversations/YYYY-MM-DD.md

algernon-review阿尔杰农审查

algernon-review

algernon-review

Constants

FSRS-4.5 Parameters

Step 1 — Fetch Due Cards