1.0.0

Changed

• Incremental filtering: typing more characters into a plain-text filter now narrows the previous result set instead of re-scanning every entry, keeping large buffers responsive while typing
• Filter tokens are lowercased once during parsing, eliminating repeated per-line strings.ToLower allocations in the match and highlight hot paths
• Trimming old entries when the buffer cap is reached now releases the dropped lines' bytes to the garbage collector immediately rather than retaining them until the backing array is reallocated
• Filter result slice is pre-sized for the common empty-query case, avoiding repeated regrowth on startup and when clearing the filter
• History modal no longer re-runs the filter on navigation, delete, or close — only when an entry is actually applied
• Per-entry footprint reduced by removing an unused matched flag

Fixed

• A trailing line written without a newline is no longer emitted prematurely and then re-emitted as a separate line once the writer completes it
• Lines longer than 64 KB were silently truncated by the underlying scanner; the limit is now 1 MB for piped stdin and effectively unbounded for tailed files