Skip to content

DEV Community

Toolkit Labs

Small, dependency-free developer tools and test datasets for teams shipping LLM agents. Honest benchmarks, CC0 where possible.

Joined on Aug 19, 2026

Aug 19

Truncated JSON from a streaming model: 21 parsers, 25 truncated cases, 12 score 0/25

#llm #ai #python #javascript

4 min read

Aug 19

I labelled 300 broken LLM tool calls - passing them straight to the schema fails 290 of 300

#ai #llm #python #agents

5 min read

Aug 19

I benchmarked 7 Python JSON parsers on 300 malformed LLM outputs — mine lost the column I cared about

#python #ai #llm #opensource

7 min read

Aug 19

I built a 300-case conformance suite for LLM JSON parsing, and it found 18 failures in my own parser

#ai #llm #python #testing

4 min read

loading...