Tools · Memory · Feb 2026
CLN → Anki
When you write a CLN document you are already classifying every piece of information. A line marked
= is a definition. A line marked ! is a warning to remember. A line marked
. is a procedure to practice. The operator is already the card type.
01/
One Authoring Act. Two Outputs.
Traditional workflow:
take notes
→
make cards
→
review cards
Three passes. Three cognitive loads.
CLN workflow:
write CLN
→
reference doc
+
Anki deck
One pass. Semantic typing happens at authoring time.
*
Self-Test: Why does CLN eliminate a separate card-authoring step?
Because operators classify information at write time.
= → definition, ! → warning, . → procedure. The card type is already there.
02/
Operator → Card Type Mapping
| Op | Semantic Role | Card Type Generated |
|---|---|---|
| = | Definition / Essence | Basic or Cloze card Front: "What is X?" · Back: the definition |
| ! | Risk / Warning | High-priority retention Tagged "critical" · higher interval penalty |
| . | Action / Procedure | Procedural card Front: "How do you do X?" · Back: the steps |
| ? | Unknown / Hypothesis | Research card "Status of this question?" · tracks resolution |
| → | Result / Conclusion | Synthesis recall card "What is the outcome of X?" · high-value |
| ~ | Tension / Trade-off | Decision card "Which side of this trade-off applies now?" |
Example — the same CLN block produces three different card types:
= Signal-source decoupling: severing of reliable correlation between signal and source
! Heuristics fire correctly but produce wrong conclusions in AI-saturated environments
→ CLN stabilizes semantic signal by encoding type at write time
. Audit your CLN document: scan
! column for unmitigated risks ##audit-stepThis block generates: 1 definition card, 1 critical warning card, 1 synthesis card, 1 procedural card. Zero extra authoring work.
*
Self-Test: What card type does a
~ line generate?
A decision card. The tension has no resolution yet — Anki forces you to revisit it until you choose a side.
03/
Gap Detection — Thinking Audit
Most Powerful Feature
Gap detection transforms CLN from a note format into a thinking audit tool. It surfaces things you wrote down but never resolved — the most valuable items to review.
Gap Pattern 1
? without following →
= open question with no conclusion
→ generates "research needed" card 🔴
Gap Pattern 2
! without following .
= risk with no mitigation action
→ generates "action required" card 🔴
Gap Pattern 3
~ without following →
= unresolved tension or trade-off
→ generates "decide" card 🟡
The gap between?and→is where your actual thinking happens. Anki can make you review that gap until it closes.
*
Self-Test: What does a
! line without a following . mean in your notes?
An unmitigated risk. You identified a danger but never wrote the response. Gap detection flags it and forces you to decide: add a mitigation or consciously accept the risk.
04/
Authoring Patterns for Better Cards
Three patterns that maximize card quality with minimal extra effort:
Pattern A — Definition Block
= concept name: definition
+ key implication 1
+ key implication 2
→ 1 definition card + N implication cards
Pattern B — Risk-Action Pair
! risk description
. mitigation step
→ "What is the response to [risk]?" card
Pattern C — Q&A Block
? open question
→ answer or resolution
→ direct Q&A card — most natural type
Using || spoilers || (v2.2) inside * lines creates self-test moments
directly in your source document — these become cloze-deletion candidates automatically:
*
Example:
The three CLN gap patterns are:
? without →, ! without ., ~ without →
*
Example:
cln2anki step 3 is:
run gap detection — pattern matching on operator sequences
05/
cln2anki: Implementation Sketch
. Input
= any
.cln file or CLN block in .md= optional deck name from document header
#deck:tag=
|| spoiler || lines → cloze candidates (v2.2). Algorithm
= step 1: parse lines →
(operator, content, indent, block) tuples= step 2: apply operator → card-type mapping
= step 3: run gap detection — pattern match operator sequences
= step 4: extract
|| ... || spans as cloze deletions= step 5: generate Anki import file
= step 6: tag cards by block header
// and document tags. Output Formats
= Anki
.apkg via genanki (Python)= AnkiConnect JSON for direct API import
= plain CSV for any spaced-repetition tool
= Markdown with
{{cloze}} syntax for Obsidian SR
*
Self-Test: What does step 4 of cln2anki do that wasn't possible in v2.1?
Extracts || spoiler || spans as cloze deletions. The v2.2 inline spoiler syntax doubles as a card-authoring primitive — the hidden text becomes the cloze answer.