commands_used_to_find_mismatches.txt

To list all problem folders:

printf '%s\n' problems/*/*/ > problem_folders.txt
sed 's/$/ | status: TODO | notes: |/' problem_folders.txt > todo.txt
split -l 30 --numeric-suffixes=1 --suffix-length=2 --additional-suffix=.txt todo.txt todo_

Prompt:

The file @problem_folders.txt lists every problem in the repo.

Each line in @problem_folders.txt looks like this:

problems/backtracking/count-unique-submultisets-with-sum-zero/ | status: TODO | notes:

Your goal is to do something for every problem in this list.
Initially, every problem has "status: TODO".

Process @problem_folders.txt top-to-bottom. For each row with status TODO:
1) Open `<path>/raw.md`. Extract constraints from the "Constraints:" section. If absent/ambiguous → set status UNSURE with a short note; continue to next row.
2) Open `<path>/solution.py`. Locate `runTests()`.
3) Compare every test input against all extracted constraints (respecting inclusive/exclusive bounds and types).
4) If any test violates constraints → set status MISMATCH and add a one-line reason in notes (e.g., `k=-1; k>=0`).
5) Otherwise set status OK. Leave notes empty.


Rules:
- Edit only the current row.
- Never reorder rows.
- Keep notes ≤120 chars.
- Do not change any file other than @problem_folders.txt 
- If a file is missing, mark ERROR and continue.

YOUR GOAL IS TO PROCESS ROWS 150 to 180. OTHER AGENST MAY PROCESS OTHER ROWS AT THE SAME TIME.