Json answer formatter
[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.
npx -y skills add cxcscmu/SkillLearnBench --skill json-answer-formatterAssembled 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
Format retrieval answers as a dictionary in /root/answer.json.
SKILL.md
0.6 KB, 125 tokens by cl100k_base, as published. Nobody here has run it
Overview
This skill defines the format for saving the answers of retrieved data.
Output Format
{
"q1": {"answer": ["value1", "value2"], "tokens": 123},
"q2": {"answer": ["single_value"], "tokens": 456}
}
Requirements
answerfield must always be a list.tokensfield must be a positive numeric value.- Store the file at
/root/answer.json. - When appending, read the existing file if it exists, update it, and write it back.