pi-read-chunks 1.0.0 → 1.0.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/README.md +1 -1
  2. package/package.json +2 -2
package/README.md CHANGED
@@ -1,6 +1,6 @@
1
1
  # pi-read-chunks
2
2
 
3
- A chunked, summarising `read` tool for large text files. Files under a size threshold are returned verbatim; larger files are split into overlapping chunks snapped to natural boundaries (function endings for code, paragraph breaks for prose), and each chunk is summarised by the active model, chaining the running summary forward as the file is consumed.
3
+ A Pi Agent extension that enhances read functionality for large text files to reduce context bloat, rot and lost in the middle issues. Pi Agent's built in 'read()' tool truncates results over a certain size. 'read-chunks()' instead splits larger files into overlapping chunks snapped to natural boundaries (function endings for code, paragraph breaks for prose), and each chunk is summarised by the active model, chaining the running summary forward as the file is consumed. Files under a configurable size threshold have contents returned verbatim.
4
4
 
5
5
  Replaces the built-in `read()` for text files. Images and other binaries still pass through to the built-in `read`. A precise query stops the scan early at the first chunk that answers it; a line-range suffix (`file.txt:2000-2089` or `file.txt:N`) returns those lines verbatim with no summarisation.
6
6
 
package/package.json CHANGED
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "name": "pi-read-chunks",
3
- "version": "1.0.0",
4
- "description": "Pi extension that replaces the built-in read() for text files with a chunked, summarising reader. Large files are split at natural boundaries, each chunk is summarised by the active model, and a query stops the scan early at the first chunk that answers it. Line ranges and binary files pass through unchanged.",
3
+ "version": "1.0.1",
4
+ "description": "A Pi Agent extension that enhances `read` functionality for large text files to reduce context bloat, rot and lost in the middle issues.",
5
5
  "keywords": [
6
6
  "pi-package",
7
7
  "pi-extension"