DavidCho1999/archaiflow-skills / bundles/construction-admin/skills/drawing-reader
drawing-reader
UnverifiedReads an architectural drawing PDF and answers questions about it: a single sheet, a full set, a redline, a record drawing, a shop drawing, or a detail. Summarises, finds a sheet, explains a detail, lists disciplines, or pulls a callout. Trigger on "read this drawing", "what is in this set", "summarize this drawing", "show me sheet X", "what is on this sheet", or a pasted path to a drawing PDF. For specifications, codes, contracts, and reports use doc-reader instead. Detects the drawing type, asks what is wanted before reading at length, and hands off to schedule-extractor, drawing-set-bookmarker, or drawing-checklist when one of those fits better. Never reports a dimension or a note it did not read off the sheet.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.4 KB |
| references/_provenance.md | markdown | 2.4 KB |
| references/drawing-types.md | markdown | 6.0 KB |
| references/read-patterns.md | markdown | 5.4 KB |
references/read-patterns.md
View on GitHub# User-Intent → Read-Pattern Mapping
When the user drops a drawing PDF and states an intent, this file maps the intent to the right handling — either direct vision read or delegation to a specialized archtmpl skill.
> **Format note**: This file is read by Claude after the user has stated their intent. The "Direct read" rows describe what to do inline; the "Delegate to" rows tell Claude to stop and recommend a specialized skill instead of doing the work here.
---
## Sheet location ("show me sheet X")
User asks: "Show me sheet A-301", "where's the egress plan", "find the door schedule sheet".
**Direct read**:
1. Inventory the PDF (page count, first-page sheet identifier).
2. Iterate page-by-page reading just the titleblock until the requested sheet is found.
3. Render that page; output a 3–5 sentence summary of the sheet's content.
4. Cite: page number + sheet number.
**Stop signals**: if 3 passes through the document don't find the sheet, tell the user the sheet isn't in this PDF.
---
## Set summary ("summarize this set")
User asks: "Summarize this set", "what's in this PDF", "give me an overview".
**Direct read**:
1. Read the cover sheet (page 1) for sheet index + project info.
2. Sample 3–5 pages across disciplines to confirm content.
3. Output: project name, set issue date, sheet count, disciplines present, key sheets per discipline.
**Delegate to `drawing-set-bookmarker`** if the user wants a *tabulated* sheet list (CSV index + bookmarked PDF).
---
## Single-sheet summary ("what's on this sheet")
User asks: "What's on sheet X", "summarize this sheet", "explain this drawing".
**Direct read**:
1. Read the entire single page.
2. Output: sheet number + title, scale(s), key drawings (plan + sections + details), general notes presence/absence, schedule presence.
3. Cite: page number + every callout / note quoted.
---
## Detail explanation ("explain this detail")
User asks: "Explain this detail", "what materials", "what's this construction".
**Direct read**:
1. Read the single detail.
2. Identify: assembly type (e.g. wall section, parapet, sill), components (with material name verbatim from the drawing), callouts referenced (sheet/detail #), key dimensions (with `low` confidence — vision is approximate).
3. Cite: every component callout's exact text from the drawing.
**Caveat**: explicitly tell the user dimensions are vision-approximated and to verify against the source PDF for any dimension that matters.
---
## Schedule extraction ("get the door schedule")
User asks: "Extract the door schedule", "give me the finish schedule as CSV", "pull the equipment schedule".
**Delegate to `schedule-extractor`**. Stop. Do not partially do the extraction inline.
If `schedule-extractor` is not installed: `npx archtmpl@latest --skill schedule-extractor`.
---
## Set tabulation / bookmarking ("bookmark this set")
User asks: "Bookmark this set", "make a sheet index from this PDF", "tabulate every sheet".
**Delegate to `drawing-set-bookmarker`**. Stop.
If not installed: `npx archtmpl@latest --skill drawing-set-bookmarker`.
---
## QA / completeness ("review this DD set")
User asks: "QA this set", "DD checklist on this", "what's missing from CD".
**Delegate to `drawing-checklist`**. Stop.
If not installed: `npx archtmpl@latest --skill drawing-checklist`.
---
## Code analysis ("what's the code analysis")
User asks: "Pull the code analysis from this set", "what occupancy / construction type / allowable area".
**Read** the cover sheet code-analysis block (typically `G-001` or `G-002`) and quote it verbatim.
If the user wants the code analysis *built from scratch* from program data, **delegate to `code-analysis-builder`**.
---
## Egress / fire / accessibility ("show me the egress")
User asks: "Show me the egress plan", "where are the fire-rated walls", "is this accessible".
**Direct read**:
1. Locate the relevant code-analysis sheet (`G-001` / `G-002` / `LS-101` for life safety).
2. Read the legend + plan diagrams.
3. Output: the egress strategy (number of stairs, travel distance, common path), fire-rated wall locations, accessibility provisions — with sheet+page citations.
4. Confidence: `medium` for plan diagrams (vision-read), `high` for verbatim notes / legend.
**Caveat**: tell the user this is a structural read, not a code-compliance verification. For compliance checking, use `ada-tracker`, `code-analysis-builder`, etc.
---
## Markup / redline aggregation ("what does this redline say")
User asks: "Tabulate the redline comments", "what does the AHJ want", "list every comment".
**Direct read**:
1. Iterate page-by-page identifying comment bubbles + cloud locations.
2. Output a CSV: `comment_id, sheet, location, comment_text, raised_by, action_required`.
3. Confidence: `high` for verbatim comment text, `medium` for inferred location.
---
## Compare versions ("addendum 1 vs addendum 2")
User asks: "Compare these two sets", "what changed in addendum 2", "diff this CD set".
**Out of scope here**. Redirect to `issue-set-comparison` when shipped, or tell the user to do a manual sheet-by-sheet read for now.
---
## Default — unrecognized intent
If the user's intent doesn't match any of the patterns above:
1. Ask the user to clarify what they want.
2. Offer 3 candidate handlings (e.g. "I can: (a) summarize the whole set, (b) find a specific sheet, (c) extract a schedule. Which?").
3. Wait for confirmation before reading.
Never read 200 pages of a set without explicit user authorization — vision reads are token-heavy.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| canvas-designanthropics/skillsCreate beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| docxanthropics/skillsUse this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pdfanthropics/skillsUse this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pptxanthropics/skillsUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions "deck," "slides," "presentation," or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |