Run3 data format validation
[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.
npx -y skills add cxcscmu/SkillLearnBench --skill run3_data_format_validationAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
What its author says it does
Copied from the file, not written here
Ensures the generated output strictly adheres to the required JSON schema.
SKILL.md
0.6 KB, 146 tokens by cl100k_base, as published. Nobody here has run it
Before finalizing the task, write the results to /root/answers.json. Verify the structure matches:
{
"q1_answer": number,
"q2_answer": number,
"q3_answer": ["stock_cusip1", "stock_cusip2", "stock_cusip3", "stock_cusip4", "stock_cusip5"],
"q4_answer": ["fund1", "fund2", "fund3"]
}
Ensure all monetary figures are represented as integers or floats consistent with the validated units (e.g., if the raw data is in thousands, the final q1_answer must be multiplied by 1,000).