artifact-type-tailored-context

To compress large code diffs and logs within a token budget, this Skill applies tiered summarization strategies that preserve function signatures while summarizing method bodies and truncates content when necessary to guarantee output stays within limits.

The best way to summarize JSON artifacts for judge evaluation is using type-specific compression strategies within a token budget, which reads raw artifacts, counts tokens, and returns compacted content with metadata using forked context for isolation.

Tiered summarization for markdown and code diffs works by applying different compression strategies based on artifact size and type, preserving critical structural elements like function signatures while condensing the surrounding content to fit token limits.

Yes, you can enforce a strict token limit when compressing large log files. The Skill guarantees output stays within specified token limits by counting tokens and applying hard truncation when necessary to prevent context window overflow.

Yes, this artifact compression approach supports markdown and JSON files. It applies specific compression strategies tailored to each artifact type, including code diffs and logs, ensuring critical information is preserved without overwhelming the context window.