paddleocr-doc-parsing

Name: paddleocr-doc-parsing
Author: freestylefly

342installs95forks

How to install paddleocr-doc-parsing?

freestylefly/canghe-skills/paddleocr-doc-parsing

$npx -y skills add freestylefly/canghe-skills --skill paddleocr-doc-parsing

Installs into the current project.

Summary

Advanced document parsing with PaddleOCR. Returns complete document structure including text, tables, formulas, charts, and layout information. The AI agent extracts relevant content based on user needs.

Files · 1

View on GitHub

SKILL.md

1	# PaddleOCR Document Parsing Skill
2
3	## When to Use This Skill
4
5	Use Document Parsing for:
6	- Documents with tables (invoices, financial reports, spreadsheets)
7	- Documents with mathematical formulas (academic papers, scientific documents)
8	- Documents with charts and diagrams
9	- Multi-column layouts (newspapers, magazines, brochures)
10	- Complex document structures requiring layout analysis
11	- Any document requiring structured understanding
12
13	Use Text Recognition instead for:
14	- Simple text-only extraction
15	- Quick OCR tasks where speed is critical
16	- Screenshots or simple images with clear text
17
18	## How to Use This Skill
19
20	⛔ MANDATORY RESTRICTIONS - DO NOT VIOLATE ⛔
21
22	1. ONLY use PaddleOCR Document Parsing API - Execute the script `python scripts/vl_caller.py`
23	2. NEVER parse documents directly - Do NOT parse documents yourself
24	3. NEVER offer alternatives - Do NOT suggest "I can try to analyze it" or similar
25	4. IF API fails - Display the error message and STOP immediately
26	5. NO fallback methods - Do NOT attempt document parsing any other way
27
28	If the script execution fails (API not configured, network error, etc.):
29	- Show the error message to the user
30	- Do NOT offer to help using your vision capabilities
31	- Do NOT ask "Would you like me to try parsing it?"
32	- Simply stop and wait for user to fix the configuration
33
34	### Basic Workflow
35
36	1. Execute document parsing:
37	```bash
38	python scripts/vl_caller.py --file-url "URL provided by user" --pretty
39	```
40	Or for local files:
41	```bash
42	python scripts/vl_caller.py --file-path "file path" --pretty
43	```
44
45	Optional: explicitly set file type:
46	```bash
47	python scripts/vl_caller.py --file-url "URL provided by user" --file-type 0 --pretty
48	```
49	- `--file-type 0`: PDF
50	- `--file-type 1`: image
51	- If omitted, the service can infer file type from input.
52
53	Default behavior: save raw JSON to a temp file:
54	- If `--output` is omitted, the script saves automatically under the system temp directory
55	- Default path pattern: `<system-temp>/paddleocr/doc-parsing/results/result_<timestamp>_<id>.json`
56	- If `--output` is provided, it overrides the default temp-file destination
57	- If `--stdout` is provided, JSON is printed to stdout and no file is saved
58	- In save mode, the script prints the absolute saved path on stderr: `Result saved to: /absolute/path/...`
59	- In default/custom save mode, read and parse the saved JSON file before responding
60	- In save mode, always tell the user the saved file path and that full raw JSON is available there
61	- Use `--stdout` only when you explicitly want to skip file persistence
62
63	2. The output JSON contains COMPLETE content with all document data:
64	- Headers, footers, page numbers
65	- Main text content
66	- Tables with structure
67	- Formulas (with LaTeX)
68	- Figures and charts
69	- Footnotes and references
70	- Seals and stamps
71	- Layout and reading order
72
73	Input type note:
74	- Supported file types depend on the model and endpoint configuration.
75	- Always follow the file type constraints documented by your endpoint API.
76
77	3. Extract what the user needs from the output JSON using these fields:
78	- Top-level `text`
79	- `result[n].markdown`
80	- `result[n].prunedResult`
81
82	### IMPORTANT: Complete Content Display
83
84	CRITICAL: You must display the COMPLETE extracted content to the user based on their needs.
85
86	- The output JSON contains ALL document content in a structured format
87	- In save mode, the raw provider result can be inspected in the saved JSON file
88	- Display the full content requested by the user, do NOT truncate or summarize
89	- If user asks for "all text", show the entire `text` field
90	- If user asks for "tables", show ALL tables in the document
91	- If user asks for "main content", filter out headers/footers but show ALL body text
92
93	What this means:
94	- DO: Display complete text, all tables, all formulas as requested
95	- DO: Present content using these fields: top-level `text`, `result[n].markdown`, and `result[n].prunedResult`
96	- DON'T: Truncate with "..." unless content is excessively long (>10,000 chars)
97	- DON'T: Summarize or provide excerpts when user asks for full content
98	- DON'T: Say "Here's a preview" when user expects complete output
99
100	Example - Correct:
101	```
102	User: "Extract all the text from this document"
103	Agent: I've parsed the complete d

People who liked this also viewed

6 picks

Type

paddleocr-doc-parsing

342installs95forks

View on GitHub

How to install paddleocr-doc-parsing?

freestylefly/canghe-skills/paddleocr-doc-parsing

$npx -y skills add freestylefly/canghe-skills --skill paddleocr-doc-parsing

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# PaddleOCR Document Parsing Skill
2
3	## When to Use This Skill
4
5	Use Document Parsing for:
6	- Documents with tables (invoices, financial reports, spreadsheets)
7	- Documents with mathematical formulas (academic papers, scientific documents)
8	- Documents with charts and diagrams
9	- Multi-column layouts (newspapers, magazines, brochures)
10	- Complex document structures requiring layout analysis
11	- Any document requiring structured understanding
12
13	Use Text Recognition instead for:
14	- Simple text-only extraction
15	- Quick OCR tasks where speed is critical
16	- Screenshots or simple images with clear text
17
18	## How to Use This Skill
19
20	⛔ MANDATORY RESTRICTIONS - DO NOT VIOLATE ⛔
21
22	1. ONLY use PaddleOCR Document Parsing API - Execute the script `python scripts/vl_caller.py`
23	2. NEVER parse documents directly - Do NOT parse documents yourself
24	3. NEVER offer alternatives - Do NOT suggest "I can try to analyze it" or similar
25	4. IF API fails - Display the error message and STOP immediately
26	5. NO fallback methods - Do NOT attempt document parsing any other way
27
28	If the script execution fails (API not configured, network error, etc.):
29	- Show the error message to the user
30	- Do NOT offer to help using your vision capabilities
31	- Do NOT ask "Would you like me to try parsing it?"
32	- Simply stop and wait for user to fix the configuration
33
34	### Basic Workflow
35
36	1. Execute document parsing:
37	```bash
38	python scripts/vl_caller.py --file-url "URL provided by user" --pretty
39	```
40	Or for local files:
41	```bash
42	python scripts/vl_caller.py --file-path "file path" --pretty
43	```
44
45	Optional: explicitly set file type:
46	```bash
47	python scripts/vl_caller.py --file-url "URL provided by user" --file-type 0 --pretty
48	```
49	- `--file-type 0`: PDF
50	- `--file-type 1`: image
51	- If omitted, the service can infer file type from input.
52
53	Default behavior: save raw JSON to a temp file:
54	- If `--output` is omitted, the script saves automatically under the system temp directory
55	- Default path pattern: `<system-temp>/paddleocr/doc-parsing/results/result_<timestamp>_<id>.json`
56	- If `--output` is provided, it overrides the default temp-file destination
57	- If `--stdout` is provided, JSON is printed to stdout and no file is saved
58	- In save mode, the script prints the absolute saved path on stderr: `Result saved to: /absolute/path/...`
59	- In default/custom save mode, read and parse the saved JSON file before responding
60	- In save mode, always tell the user the saved file path and that full raw JSON is available there
61	- Use `--stdout` only when you explicitly want to skip file persistence
62
63	2. The output JSON contains COMPLETE content with all document data:
64	- Headers, footers, page numbers
65	- Main text content
66	- Tables with structure
67	- Formulas (with LaTeX)
68	- Figures and charts
69	- Footnotes and references
70	- Seals and stamps
71	- Layout and reading order
72
73	Input type note:
74	- Supported file types depend on the model and endpoint configuration.
75	- Always follow the file type constraints documented by your endpoint API.
76
77	3. Extract what the user needs from the output JSON using these fields:
78	- Top-level `text`
79	- `result[n].markdown`
80	- `result[n].prunedResult`
81
82	### IMPORTANT: Complete Content Display
83
84	CRITICAL: You must display the COMPLETE extracted content to the user based on their needs.
85
86	- The output JSON contains ALL document content in a structured format
87	- In save mode, the raw provider result can be inspected in the saved JSON file
88	- Display the full content requested by the user, do NOT truncate or summarize
89	- If user asks for "all text", show the entire `text` field
90	- If user asks for "tables", show ALL tables in the document
91	- If user asks for "main content", filter out headers/footers but show ALL body text
92
93	What this means:
94	- DO: Display complete text, all tables, all formulas as requested
95	- DO: Present content using these fields: top-level `text`, `result[n].markdown`, and `result[n].prunedResult`
96	- DON'T: Truncate with "..." unless content is excessively long (>10,000 chars)
97	- DON'T: Summarize or provide excerpts when user asks for full content
98	- DON'T: Say "Here's a preview" when user expects complete output
99
100	Example - Correct:
101	```
102	User: "Extract all the text from this document"
103	Agent: I've parsed the complete d

People who liked this also viewed

6 picks

Type