improve: expand compare_and_contrast and create_flash_cards; apply cognitive science to flashcards#2126
Open
OdinKral wants to merge 3 commits into
Open
improve: expand compare_and_contrast and create_flash_cards; apply cognitive science to flashcards#2126OdinKral wants to merge 3 commits into
OdinKral wants to merge 3 commits into
Conversation
93 patterns had no standardized # INPUT section, making the input boundary ambiguous. Three variants existed: - 7 patterns used ## INPUT (wrong heading level) → upgraded to # INPUT - 8 patterns had bare "INPUT:" text with no heading → added # INPUT header - 78 patterns had no input section at all → appended # INPUT + INPUT: All changes are structural only — no content was altered. The # INPUT section is the standard marker used by Fabric's CLI and pattern tooling to delimit where user input is injected. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
compare_and_contrast (14 lines → 20): was "Please be brief. Compare and contrast the list of items." with no guidance on how to pick dimensions or what makes a useful comparison. Expanded with dimension-selection step, cell-length guidance, and a Key Takeaways section after the table. create_flash_cards (14 lines → 22): had no guidance on card count, no format spec beyond "Markdown", and an unnecessary output section. Expanded with numbered steps, card count target, Q:/A: format with length constraints, and a reminder to test understanding over trivia. raw_query intentionally left as-is — it's a minimal pass-through pattern by design. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Expands the pattern with learning principles from active recall and chunking research (Barbara Oakley, "A Mind for Numbers"): - Retrieval vs recognition: questions must force generation, not pattern-matching. Includes a wrong/right example in the pattern. - One idea per card: explicitly enforced — compound questions split. - Concrete anchors: abstract answers must end with a concrete example. - Three question types: Definition, Mechanism/Why-How, Application/When — all three used where the material supports it. - Synthesis cards: 2–4 cards per deck connecting concepts into chunks. Output format updated: Q:/A: labels (cleaner than **Q:**/**A:**), 40-word answer limit (up from 32 to accommodate examples), and a loose card ordering (Definition → Mechanism → Application → Synthesis) that mirrors how understanding builds. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
OdinKral
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Two stub patterns expanded with real structure.
create_flash_cardsadditionally updated with cognitive science principles from active recall research.compare_and_contrastBefore: "Please be brief. Compare and contrast the list of items." with no dimension-selection guidance.
After: Steps for picking meaningful dimensions (not trivial similarities), cell-length guidance, and a Key Takeaways section to surface non-obvious patterns.
create_flash_cardsBefore: 4 bullets, no card count, no format spec.
After (initial): Numbered steps, card count target, Q:/A: format.
After (cognitive science pass): Added principles from active recall research:
(e.g. ...).raw_querywas also flagged by the audit script but intentionally left as-is — it's a minimal pass-through by design.Test plan
fabric -p compare_and_contrast "Python vs JavaScript vs Go"returns a table with meaningful dimensions and a Key Takeaways sectionfabric -p create_flash_cardson a short article returns Q:/A: cards across all three question types, with concrete examples on abstract answers, and 2–4 synthesis cards at the end🤖 Generated with Claude Code