@jarkkojs/readseek 0.6.6 → 0.6.8

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 +12 -4
  2. package/package.json +5 -5
package/README.md CHANGED
@@ -19,7 +19,10 @@ npm install -g @jarkkojs/readseek
19
19
  ```
20
20
 
21
21
  Prebuilt binaries are available for macOS ARM64, Linux ARM64 and x64, and
22
- Windows x64. The Linux binaries are statically linked with musl.
22
+ Windows x64. The Linux binaries are static glibc PIE executables.
23
+
24
+ Local builds require CMake, Clang/libclang, and a C++ compiler because image
25
+ inference uses `llama-cpp-2`.
23
26
 
24
27
  ## Pi extension
25
28
 
@@ -63,21 +66,23 @@ printf '%s\n' 'fn main() {}' | readseek identify stdin:scratch.rs:1 --column 4
63
66
  ## Images and PDFs
64
67
 
65
68
  `detect` reports image metadata and PDF page counts. `read` returns bounded
66
- base64 images by default; use `--image` for one local analysis mode:
69
+ base64 images by default; use `--image` for a local analysis mode:
67
70
 
68
71
  ```sh
69
72
  readseek read photo.jpg # default: bounded base64 image
70
73
  readseek read photo.jpg --image caption # detailed natural-language caption
71
74
  readseek read photo.jpg --image objects # object labels + bounding boxes
72
75
  readseek read photo.jpg --image ocr # extracted text
76
+ readseek read photo.jpg --image all # caption, objects, and OCR in one pass
73
77
  ```
74
78
 
75
79
  PDF reads return page-tagged Markdown and page-associated embedded images. The
76
80
  same mode applies to each embedded image. Line/hash suffixes, `--end`, `--limit`,
77
81
  and `--language` do not apply to visual files.
78
82
 
79
- Vision models download lazily into the user cache and run on the CPU. Captioning
80
- can take substantial time.
83
+ The Qwen3-VL GGUF model and multimodal projector download lazily into the user
84
+ cache and run through `llama-cpp-2` on the CPU. Captioning can take substantial
85
+ time.
81
86
 
82
87
  ## Cache
83
88
 
@@ -100,3 +105,6 @@ Pass `--help` to any command for command-specific usage.
100
105
 
101
106
  `readseek` is licensed under `LGPL-2.1-or-later`. The JavaScript npm wrapper
102
107
  is licensed under `Apache-2.0`.
108
+
109
+ The downloaded `Qwen/Qwen3-VL-2B-Instruct-GGUF` model is licensed under
110
+ `Apache-2.0`.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@jarkkojs/readseek",
3
- "version": "0.6.6",
3
+ "version": "0.6.8",
4
4
  "description": "A structural read command",
5
5
  "license": "Apache-2.0 AND LGPL-2.1-or-later",
6
6
  "homepage": "https://github.com/jarkkojs/readseek#readme",
@@ -21,10 +21,10 @@
21
21
  "LICENSE-LGPL-2.1"
22
22
  ],
23
23
  "optionalDependencies": {
24
- "@jarkkojs/readseek-darwin-arm64": "0.6.6",
25
- "@jarkkojs/readseek-linux-arm64": "0.6.6",
26
- "@jarkkojs/readseek-linux-x64": "0.6.6",
27
- "@jarkkojs/readseek-win32-x64": "0.6.6"
24
+ "@jarkkojs/readseek-darwin-arm64": "0.6.8",
25
+ "@jarkkojs/readseek-linux-arm64": "0.6.8",
26
+ "@jarkkojs/readseek-linux-x64": "0.6.8",
27
+ "@jarkkojs/readseek-win32-x64": "0.6.8"
28
28
  },
29
29
  "publishConfig": {
30
30
  "access": "public"