sheet-splitter

Name: sheet-splitter
Author: dleerdefi

24stars5forks

How to install sheet-splitter?

dleerdefi/claude-code-construction/sheet-splitter

$npx -y skills add dleerdefi/claude-code-construction --skill sheet-splitter

Installs into the current project.

Summary

Split a bound drawing set PDF into individual sheet PDFs. Extracts sheet numbers from title blocks via vision, renames files. Triggers: 'split drawings', 'break up the drawing set', 'separate sheets'. Run after /project-setup.

Files · 1

View on GitHub

SKILL.md

1	# Sheet Splitter
2
3	Splits a bound drawing set PDF into individual page PDFs. Mirrors what spec-splitter does for specifications.
4
5	## Pipeline Position
6	Run after `/project-setup` identifies bound drawing sets. Produces split PDFs and `sheet_index.yaml` consumed by `/schedule-extractor` and all drawing analysis skills.
7
8	This is valuable for:
9	- Project teams: Navigate drawings by sheet number instead of scrolling a multi-page PDF
10	- Other skills: `sheet-index-builder`, `drawing-reader`, and all drawing analysis skills work better with individual sheet files
11	- AgentCM: Split files become the basis for per-sheet structured data
12
13	## Workflow
14
15	```
16	Sheet Split Progress:
17	- [ ] Step 1: Check if drawings are already split
18	- [ ] Step 2: Locate the bound drawing set
19	- [ ] Step 3: Split into individual page PDFs
20	- [ ] Step 4: Extract sheet numbers from title blocks
21	- [ ] Step 5: Rename files with sheet number + title
22	- [ ] Step 6: Update project context
23	- [ ] Step 7: Write graph entry (AgentCM only)
24	```
25
26	### Step 1: Check if Drawings Are Already Split
27
28	Look for individual sheet PDFs. Drawings are already split if:
29	- AgentCM: `.construction/index/sheet_index.yaml` exists with individual sheet entries and file paths point to split PDFs
30	- Multiple PDFs exist with sheet number patterns in filenames (e.g., `A-1.1 - FLOOR PLAN.pdf`)
31	- A `sheet_index.yaml` exists in the drawings directory
32
33	If already split, report the count and skip. If AgentCM has the sheet index but files haven't been physically split on disk, offer to split using the existing metadata for naming.
34
35	### Step 2: Find ALL Drawing PDFs
36
37	Search the project directory for ALL PDFs that are drawings. Many projects have multiple drawing PDFs:
38	- Multi-part sets: Part1.pdf, Part2.pdf, Part3.pdf (split by file size)
39	- Discipline sets: Architectural.pdf, Civil.pdf, Electrical.pdf, Mechanical.pdf
40	- Single combined: one large PDF with all sheets
41
42	Search in:
43	- Folders named `drawings/`, `plans/`, `01 - Drawings/`, or similar (case-insensitive)
44	- The project root (some projects have no folder structure)
45	- Look for PDFs > 2MB with keywords: "plan", "drawing", "bid set", "dwg", discipline names
46
47	Process EACH PDF found. All split pages go to the same `sheets/` output directory.
48
49	If the user specifies a single file (`/sheet-splitter path/to/specific.pdf`), process only that file.
50
51	### Step 3: Split Pages
52
53	Run the split script on each drawing PDF found:
54
55	```bash
56	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/scripts/split_drawing_set.py \
57	"{drawing_set.pdf}" \
58	--output-dir "{drawings_directory}/sheets"
59	```
60
61	The script splits each page into its own PDF: `page_001.pdf`, `page_002.pdf`, etc. It does NOT attempt to read sheet numbers — that's unreliable across different title block formats.
62
63	### Step 4: Identify Sheets via Vision
64
65	After splitting, read each page to identify sheet numbers and titles:
66
67	1. Rasterize the title block region of each page:
68	```bash
69	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/../../scripts/pdf/rasterize_page.py page_NNN.pdf 1 --dpi 200 --output tb.png
70	```
71
72	2. Use vision on the rasterized title block: "Read the title block on this construction drawing. What is the sheet number and sheet title?"
73
74	3. Rename the file: `page_001.pdf` → `A-1.1 - FLOOR PLAN.pdf`
75
76	4. Update `sheet_index.yaml` with the identified sheet number, title, and discipline.
77
78	Title blocks vary significantly across firms — they can be on the east side, south side, or bottom-center. Some use VA standard forms, some use firm-specific formats. Vision handles all these variations.
79
80	### Step 5: Write Sheet Index
81
82	After identifying all sheets, write `.construction/index/sheet_index.yaml`.
83
84	### Output
85
86	```
87	sheets/
88	A-1.1 - FLOOR PLAN.pdf ← renamed by Claude via vision
89	A-2.1 - ELEVATIONS.pdf
90	S-1.1 - FOUNDATION PLAN.pdf
91	page_006.pdf ← could not read title block
92	...
93	sheet_index.yaml
94	```
95
96	### Step 6: Update Project Context
97
98	After splitting and identifying, update `.construction/project_context.yaml` with:
99	- `documents.drawing_count`: number of sheets
100	- `documents.disciplines`: list of unique discipline prefixes found
101
102	### Step 7: Write Graph Entry (AgentCM only)
103
104	If `.construction/` directory exists, write a graph entry:
105
106	```bash
107	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/../../scripts/graph/write_finding.py \
108	--type "drawings_split" \
109	--title "Drawing set split: {N} sheets from {source_pdf}" \
110	--data '{"sheet_count": N, "source_pdf": "...", "identified_count": M, "unidentified_count": K, "disc

People who liked this also viewed

6 picks

Type

sheet-splitter

24stars5forks

View on GitHub

How to install sheet-splitter?

dleerdefi/claude-code-construction/sheet-splitter

$npx -y skills add dleerdefi/claude-code-construction --skill sheet-splitter

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Sheet Splitter
2
3	Splits a bound drawing set PDF into individual page PDFs. Mirrors what spec-splitter does for specifications.
4
5	## Pipeline Position
6	Run after `/project-setup` identifies bound drawing sets. Produces split PDFs and `sheet_index.yaml` consumed by `/schedule-extractor` and all drawing analysis skills.
7
8	This is valuable for:
9	- Project teams: Navigate drawings by sheet number instead of scrolling a multi-page PDF
10	- Other skills: `sheet-index-builder`, `drawing-reader`, and all drawing analysis skills work better with individual sheet files
11	- AgentCM: Split files become the basis for per-sheet structured data
12
13	## Workflow
14
15	```
16	Sheet Split Progress:
17	- [ ] Step 1: Check if drawings are already split
18	- [ ] Step 2: Locate the bound drawing set
19	- [ ] Step 3: Split into individual page PDFs
20	- [ ] Step 4: Extract sheet numbers from title blocks
21	- [ ] Step 5: Rename files with sheet number + title
22	- [ ] Step 6: Update project context
23	- [ ] Step 7: Write graph entry (AgentCM only)
24	```
25
26	### Step 1: Check if Drawings Are Already Split
27
28	Look for individual sheet PDFs. Drawings are already split if:
29	- AgentCM: `.construction/index/sheet_index.yaml` exists with individual sheet entries and file paths point to split PDFs
30	- Multiple PDFs exist with sheet number patterns in filenames (e.g., `A-1.1 - FLOOR PLAN.pdf`)
31	- A `sheet_index.yaml` exists in the drawings directory
32
33	If already split, report the count and skip. If AgentCM has the sheet index but files haven't been physically split on disk, offer to split using the existing metadata for naming.
34
35	### Step 2: Find ALL Drawing PDFs
36
37	Search the project directory for ALL PDFs that are drawings. Many projects have multiple drawing PDFs:
38	- Multi-part sets: Part1.pdf, Part2.pdf, Part3.pdf (split by file size)
39	- Discipline sets: Architectural.pdf, Civil.pdf, Electrical.pdf, Mechanical.pdf
40	- Single combined: one large PDF with all sheets
41
42	Search in:
43	- Folders named `drawings/`, `plans/`, `01 - Drawings/`, or similar (case-insensitive)
44	- The project root (some projects have no folder structure)
45	- Look for PDFs > 2MB with keywords: "plan", "drawing", "bid set", "dwg", discipline names
46
47	Process EACH PDF found. All split pages go to the same `sheets/` output directory.
48
49	If the user specifies a single file (`/sheet-splitter path/to/specific.pdf`), process only that file.
50
51	### Step 3: Split Pages
52
53	Run the split script on each drawing PDF found:
54
55	```bash
56	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/scripts/split_drawing_set.py \
57	"{drawing_set.pdf}" \
58	--output-dir "{drawings_directory}/sheets"
59	```
60
61	The script splits each page into its own PDF: `page_001.pdf`, `page_002.pdf`, etc. It does NOT attempt to read sheet numbers — that's unreliable across different title block formats.
62
63	### Step 4: Identify Sheets via Vision
64
65	After splitting, read each page to identify sheet numbers and titles:
66
67	1. Rasterize the title block region of each page:
68	```bash
69	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/../../scripts/pdf/rasterize_page.py page_NNN.pdf 1 --dpi 200 --output tb.png
70	```
71
72	2. Use vision on the rasterized title block: "Read the title block on this construction drawing. What is the sheet number and sheet title?"
73
74	3. Rename the file: `page_001.pdf` → `A-1.1 - FLOOR PLAN.pdf`
75
76	4. Update `sheet_index.yaml` with the identified sheet number, title, and discipline.
77
78	Title blocks vary significantly across firms — they can be on the east side, south side, or bottom-center. Some use VA standard forms, some use firm-specific formats. Vision handles all these variations.
79
80	### Step 5: Write Sheet Index
81
82	After identifying all sheets, write `.construction/index/sheet_index.yaml`.
83
84	### Output
85
86	```
87	sheets/
88	A-1.1 - FLOOR PLAN.pdf ← renamed by Claude via vision
89	A-2.1 - ELEVATIONS.pdf
90	S-1.1 - FOUNDATION PLAN.pdf
91	page_006.pdf ← could not read title block
92	...
93	sheet_index.yaml
94	```
95
96	### Step 6: Update Project Context
97
98	After splitting and identifying, update `.construction/project_context.yaml` with:
99	- `documents.drawing_count`: number of sheets
100	- `documents.disciplines`: list of unique discipline prefixes found
101
102	### Step 7: Write Graph Entry (AgentCM only)
103
104	If `.construction/` directory exists, write a graph entry:
105
106	```bash
107	${CLAUDE_SKILL_DIR}/../../bin/construction-python ${CLAUDE_SKILL_DIR}/../../scripts/graph/write_finding.py \
108	--type "drawings_split" \
109	--title "Drawing set split: {N} sheets from {source_pdf}" \
110	--data '{"sheet_count": N, "source_pdf": "...", "identified_count": M, "unidentified_count": K, "disc

People who liked this also viewed

6 picks

Type