categorize-leetcode

name	categorize-leetcode
description	Categorize new LeetCode problem markdown files with difficulty and topic tags, then update STUDY_INDEX.md and the interview techniques catalog. Use when the user adds a new coding problem, asks to categorize or tag problems, or asks to update the study index or techniques guide.

Categorize LeetCode Problems

When to use

Apply this skill when:

• A new {id}_{title}.md problem file is added to the repo root
• The user asks to categorize, tag, or organize LeetCode problems
• The user asks to update STUDY_INDEX.md
• The user asks to update the interview techniques guide or problem → technique mapping

Workflow

1. Identify the new or changed problem file(s) in the repo root (pattern: \d{4}_.+\.md).
2. Tag the file(s):

   python scripts/tag_problems.py <filename>.md
   

   Or tag all problems:

   python scripts/tag_problems.py --all
   

3. Regenerate the study index:

   python scripts/generate_index.py
   

4. Refresh the problem → technique mapping and appendix stub:

   python scripts/generate_technique_mapping.py
   

5. If the techniques appendix changed, splice it into docs/interview_techniques.md:
   • Replace the content after <!-- APPENDIX_PLACEHOLDER -->, or
   • Replace everything from ## Appendix A: Problems by Technique onward with scripts/_techniques_appendix.md
6. Report the assigned difficulty, topic tags, and primary technique to the user.

What the scripts do

• scripts/tag_problems.py: Fetches official difficulty and topic tags from the LeetCode API (cached in scripts/lc_metadata.json), then updates each markdown file:
  • Adds Easy/Medium/Hard after the title if missing
  • Removes legacy ### Tags sections if present
  • Adds or updates a **Tags** section with official LeetCode topic tags
  • Preserves personal flags: revisit, unsolved, non-intuitive, backwards
• scripts/generate_index.py: Reads all tagged problem files and writes STUDY_INDEX.md, grouped by topic then by difficulty (Easy → Medium → Hard → Unknown). Also:
  • Injects per-topic Must-know summaries from scripts/topic_techniques.json
  • Wraps link targets in angle brackets when filenames contain ( or ) (e.g. 0380_Insert_Delete_GetRandom_O(1).md)
  • Adds a pointer to docs/interview_techniques.md at the top of the index
• scripts/generate_technique_mapping.py: Assigns each problem a primary interview technique from its tags (with manual overrides in the script). Writes:
  • scripts/problem_techniques.json — machine-readable mapping
  • scripts/_techniques_appendix.md — appendix content for docs/interview_techniques.md

Study guide files

File	Role	Editable by hand?
STUDY_INDEX.md	Problem index by topic/difficulty	No — regenerate with generate_index.py
docs/interview_techniques.md	Must-know patterns + appendices	Yes — core pattern sections are hand-curated; appendices are regenerated
scripts/topic_techniques.json	Per-topic must-know bullets for STUDY_INDEX	Yes — edit when adding topics or refining summaries
scripts/problem_techniques.json	Problem → primary technique mapping	No — regenerate with generate_technique_mapping.py

Maintenance notes

Adding a new problem

Run steps 1–4 in the workflow above. Splice the appendix (step 5) only when the new problem should appear in Appendix A/B of the techniques doc.

If the new problem is a canonical example for a pattern, also add it manually to the relevant section in docs/interview_techniques.md (the ★ Must-do or Also practice lists).

Editing per-topic must-know summaries

Edit scripts/topic_techniques.json, then rerun:

python scripts/generate_index.py

Each entry has:

• techniques — 2–4 bullet phrases shown in STUDY_INDEX
• doc_anchor — anchor in docs/interview_techniques.md (e.g. linked-list-patterns)

Improving technique assignments

Primary technique mapping lives in scripts/generate_technique_mapping.py:

• TAG_TO_TECHNIQUE — priority-ordered tag → technique name
• OVERRIDES — problem number → technique for exceptions

After editing, rerun generate_technique_mapping.py and splice the appendix.

Link gotchas

• Filenames with parentheses must use angle-bracket links: [text](<0380_Insert_Delete_GetRandom_O(1).md>)
• generate_index.py handles this automatically for STUDY_INDEX
• From docs/interview_techniques.md, use [text](<../filename.md>) for parenthesis filenames
• One filename has a space: 0027_Remove Element.md — link as [27. Remove Element](../0027_Remove Element.md)

Do not edit by hand

• STUDY_INDEX.md — overwritten by generate_index.py
• scripts/_techniques_appendix.md — build artifact; splice into techniques doc instead of editing in place
• scripts/problem_techniques.json — regenerated from tags/overrides

File format expectations

Problem files should have a title line like:

### 1. Two Sum

or plain:

1. Two Sum

After tagging, files will include difficulty and tags:

### 1. Two Sum
Easy
...
**Tags**
- Array
- Hash Table

Notes

• Slug lookup uses the 4-digit problem ID from the filename via scripts/lc_problem_slugs.json.
• Re-running tag scripts is safe; metadata is cached and files are only rewritten when changed.
• Personal study flags in **Tags** are kept; they are excluded from topic grouping in STUDY_INDEX.md.
• Navigation: README.md links to STUDY_INDEX.md and docs/interview_techniques.md.