vidistill 0.2.4 → 0.3.0
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.
- package/README.md +8 -26
- package/dist/index.js +1032 -190
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -50,31 +50,6 @@ vidistill ./demo.mp4 -o ./notes/
|
|
|
50
50
|
vidistill ./lecture.mp4 --lang zh
|
|
51
51
|
```
|
|
52
52
|
|
|
53
|
-
### Extract
|
|
54
|
-
|
|
55
|
-
Pull specific data from a previously processed video or re-run a targeted pass on a video file.
|
|
56
|
-
|
|
57
|
-
```
|
|
58
|
-
vidistill extract <type> <source>
|
|
59
|
-
```
|
|
60
|
-
|
|
61
|
-
**Arguments:**
|
|
62
|
-
|
|
63
|
-
- `type` — what to extract: `code`, `links`, `people`, `transcript`, or `commands`
|
|
64
|
-
- `source` — path to a vidistill output directory or a video/audio file
|
|
65
|
-
|
|
66
|
-
**Examples:**
|
|
67
|
-
|
|
68
|
-
```bash
|
|
69
|
-
# Extract code from existing output (no API calls)
|
|
70
|
-
vidistill extract code ./vidistill-output/my-video/
|
|
71
|
-
|
|
72
|
-
# Extract links from a video file (runs targeted pipeline)
|
|
73
|
-
vidistill extract links ./lecture.mp4
|
|
74
|
-
```
|
|
75
|
-
|
|
76
|
-
When pointed at an output directory, extract reads from already-generated files with zero API calls. When pointed at a video file, it runs a minimal pipeline with only the passes needed for the requested data type.
|
|
77
|
-
|
|
78
53
|
## API Key
|
|
79
54
|
|
|
80
55
|
vidistill needs a Gemini API key. It checks these sources in order:
|
|
@@ -103,12 +78,19 @@ vidistill-output/my-video/
|
|
|
103
78
|
├── action-items.md # tasks and follow-ups
|
|
104
79
|
├── insights.md # implicit signals and analysis
|
|
105
80
|
├── links.md # all URLs mentioned
|
|
81
|
+
├── prereqs.md # prerequisite knowledge (when detected)
|
|
82
|
+
├── timeline.html # interactive visual timeline
|
|
106
83
|
├── metadata.json # processing metadata
|
|
84
|
+
├── progress.json # resume checkpoint (during processing)
|
|
107
85
|
└── raw/ # raw pass outputs
|
|
108
86
|
```
|
|
109
87
|
|
|
110
88
|
Which files are generated depends on the video content — a coding tutorial gets `code/`, a meeting gets `people.md` and `action-items.md`, etc.
|
|
111
89
|
|
|
90
|
+
### Resume
|
|
91
|
+
|
|
92
|
+
If a run is interrupted (Ctrl+C), progress is saved automatically. Re-running the same command detects the incomplete run and offers to resume from where it left off.
|
|
93
|
+
|
|
112
94
|
## How It Works
|
|
113
95
|
|
|
114
96
|
Supported video formats: MP4, MOV, WebM, MKV, AVI, MPEG, FLV, WMV, 3GPP. Supported audio formats: MP3, AAC, WAV, FLAC, OGG, M4A.
|
|
@@ -118,7 +100,7 @@ Supported video formats: MP4, MOV, WebM, MKV, AVI, MPEG, FLV, WMV, 3GPP. Support
|
|
|
118
100
|
3. **Pass 1** — transcript extraction with speaker identification
|
|
119
101
|
4. **Pass 2** — visual content extraction (screen states, diagrams, slides)
|
|
120
102
|
5. **Pass 3** — specialist passes based on video type:
|
|
121
|
-
- 3c: chat and links (live streams) — per segment
|
|
103
|
+
- 3c: chat and links (live streams) — per segment, runs 3x with consensus voting
|
|
122
104
|
- 3d: implicit signals (all types) — per segment
|
|
123
105
|
- 3b: people and social dynamics (meetings) — whole video
|
|
124
106
|
- 3a: code reconstruction (coding videos) — whole video, runs 3x with consensus voting and validation
|