|
1
|
# Quick Start |
|
2
|
|
|
3
|
## Analyze a single video |
|
4
|
|
|
5
|
```bash |
|
6
|
planopticon analyze -i meeting.mp4 -o ./output |
|
7
|
``` |
|
8
|
|
|
9
|
This runs the full pipeline: |
|
10
|
|
|
11
|
1. Extracts video frames (smart sampling, change detection) |
|
12
|
2. Extracts and transcribes audio |
|
13
|
3. Detects and analyzes diagrams, charts, whiteboards |
|
14
|
4. Builds a knowledge graph of entities and relationships |
|
15
|
5. Extracts key points and action items |
|
16
|
6. Generates markdown, HTML, and PDF reports |
|
17
|
7. Outputs a `manifest.json` with everything |
|
18
|
|
|
19
|
## Processing depth |
|
20
|
|
|
21
|
```bash |
|
22
|
# Quick scan — transcription + key points only |
|
23
|
planopticon analyze -i video.mp4 -o ./out --depth basic |
|
24
|
|
|
25
|
# Standard — includes diagram extraction (default) |
|
26
|
planopticon analyze -i video.mp4 -o ./out --depth standard |
|
27
|
|
|
28
|
# Deep — more frames analyzed, richer extraction |
|
29
|
planopticon analyze -i video.mp4 -o ./out --depth comprehensive |
|
30
|
``` |
|
31
|
|
|
32
|
## Choose a provider |
|
33
|
|
|
34
|
```bash |
|
35
|
# Auto-detect best available (default) |
|
36
|
planopticon analyze -i video.mp4 -o ./out |
|
37
|
|
|
38
|
# Force a specific provider |
|
39
|
planopticon analyze -i video.mp4 -o ./out --provider openai |
|
40
|
|
|
41
|
# Use Ollama for fully offline processing (no API keys needed) |
|
42
|
planopticon analyze -i video.mp4 -o ./out --provider ollama |
|
43
|
|
|
44
|
# Override specific models |
|
45
|
planopticon analyze -i video.mp4 -o ./out \ |
|
46
|
--vision-model gpt-4o \ |
|
47
|
--chat-model claude-sonnet-4-5-20250929 |
|
48
|
``` |
|
49
|
|
|
50
|
## Batch processing |
|
51
|
|
|
52
|
```bash |
|
53
|
# Process all videos in a folder |
|
54
|
planopticon batch -i ./recordings -o ./output |
|
55
|
|
|
56
|
# Custom file patterns |
|
57
|
planopticon batch -i ./recordings -o ./output --pattern "*.mp4,*.mov" |
|
58
|
|
|
59
|
# With a title for the batch report |
|
60
|
planopticon batch -i ./recordings -o ./output --title "Q4 Sprint Reviews" |
|
61
|
``` |
|
62
|
|
|
63
|
Batch mode produces per-video outputs plus: |
|
64
|
|
|
65
|
- Merged knowledge graph across all videos |
|
66
|
- Batch summary with aggregated action items |
|
67
|
- Cross-referenced entities |
|
68
|
|
|
69
|
## Ingest documents |
|
70
|
|
|
71
|
Build a knowledge graph from documents, notes, or any text content: |
|
72
|
|
|
73
|
```bash |
|
74
|
# Ingest a single file |
|
75
|
planopticon ingest ./meeting-notes.md --output ./kb |
|
76
|
|
|
77
|
# Ingest a directory recursively |
|
78
|
planopticon ingest ./docs/ --output ./kb --recursive |
|
79
|
|
|
80
|
# Ingest from a URL |
|
81
|
planopticon ingest "https://www.youtube.com/watch?v=example" --output ./kb |
|
82
|
``` |
|
83
|
|
|
84
|
## Companion REPL |
|
85
|
|
|
86
|
Chat with your knowledge base interactively: |
|
87
|
|
|
88
|
```bash |
|
89
|
# Start the companion |
|
90
|
planopticon companion --kb ./kb |
|
91
|
|
|
92
|
# Use a specific provider |
|
93
|
planopticon companion --kb ./kb --provider anthropic |
|
94
|
``` |
|
95
|
|
|
96
|
The companion understands your knowledge graph and can answer questions, find connections, and summarize topics conversationally. |
|
97
|
|
|
98
|
## Planning agent |
|
99
|
|
|
100
|
Run the planning agent for adaptive, goal-directed analysis: |
|
101
|
|
|
102
|
```bash |
|
103
|
# Interactive mode — the agent asks before each action |
|
104
|
planopticon agent --kb ./kb --interactive |
|
105
|
|
|
106
|
# Non-interactive with export |
|
107
|
planopticon agent --kb ./kb --export ./plan.md |
|
108
|
``` |
|
109
|
|
|
110
|
## Query the knowledge graph |
|
111
|
|
|
112
|
Query your knowledge graph directly without an AI provider: |
|
113
|
|
|
114
|
```bash |
|
115
|
# Show graph stats (entity/relationship counts) |
|
116
|
planopticon query stats |
|
117
|
|
|
118
|
# List entities by type |
|
119
|
planopticon query "entities --type technology" |
|
120
|
|
|
121
|
# Find neighbors of an entity |
|
122
|
planopticon query "neighbors Alice" |
|
123
|
|
|
124
|
# Natural language query (requires API key) |
|
125
|
planopticon query "What technologies were discussed?" |
|
126
|
|
|
127
|
# Interactive REPL |
|
128
|
planopticon query -I |
|
129
|
``` |
|
130
|
|
|
131
|
## Export |
|
132
|
|
|
133
|
Export your knowledge base to various formats: |
|
134
|
|
|
135
|
```bash |
|
136
|
# Export to Markdown files |
|
137
|
planopticon export markdown --input ./kb --output ./docs |
|
138
|
|
|
139
|
# Export to an Obsidian vault |
|
140
|
planopticon export obsidian --input ./kb --output ~/Obsidian/PlanOpticon |
|
141
|
|
|
142
|
# Export to Notion |
|
143
|
planopticon export notion --input ./kb --parent-page abc123 |
|
144
|
|
|
145
|
# Export as exchange format (portable JSON) |
|
146
|
planopticon export exchange --input ./kb --output ./export.json |
|
147
|
``` |
|
148
|
|
|
149
|
## Discover available models |
|
150
|
|
|
151
|
```bash |
|
152
|
planopticon list-models |
|
153
|
``` |
|
154
|
|
|
155
|
Shows all models from configured providers with their capabilities (vision, chat, transcription). |
|
156
|
|
|
157
|
## Output structure |
|
158
|
|
|
159
|
After processing, your output directory looks like: |
|
160
|
|
|
161
|
``` |
|
162
|
output/ |
|
163
|
├── manifest.json # Single source of truth |
|
164
|
├── transcript/ |
|
165
|
│ ├── transcript.json # Full transcript with segments |
|
166
|
│ ├── transcript.txt # Plain text |
|
167
|
│ └── transcript.srt # Subtitles |
|
168
|
├── frames/ # Extracted video frames |
|
169
|
├── diagrams/ # Detected diagrams + mermaid/SVG/PNG |
|
170
|
├── captures/ # Screengrab fallbacks |
|
171
|
└── results/ |
|
172
|
├── analysis.md # Markdown report |
|
173
|
├── analysis.html # HTML report |
|
174
|
├── analysis.pdf # PDF report |
|
175
|
├── knowledge_graph.json |
|
176
|
├── key_points.json |
|
177
|
└── action_items.json |
|
178
|
``` |
|
179
|
|