Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mikeOnBreeze/cc-crossbeam --skill adu-pdf-extractiongit clone --depth 1 https://github.com/mikeOnBreeze/cc-crossbeamWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mikeonbreeze/cc-crossbeam/adu-pdf-extraction)<a href="https://agentmods.dev/skills/mikeonbreeze/cc-crossbeam/adu-pdf-extraction"><img src="https://agentmods.dev/badge/skills/mikeonbreeze/cc-crossbeam/adu-pdf-extraction/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mikeonbreeze/cc-crossbeam/adu-pdf-extraction"><img src="https://agentmods.dev/badge/skills/mikeonbreeze/cc-crossbeam/adu-pdf-extraction.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00096 | $0.04734 |
| Opus 5 | $0.00048 | $0.02367 |
| Sonnet 5 | $0.00019 | $0.00947 |
| Haiku 4.5 | $0.00010 | $0.00473 |
Grade A, and why
adu-pdf-extraction scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 441 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Construction PDF Binder Extraction
Purpose
Extract multi-page construction plan PDF binders into a vision-first structure that enables an AI agent to efficiently navigate, reference, and respond to specific pages and drawing zones within the plans.
Construction PDFs are uniquely challenging because:
- Single PDF pages often contain multiple sub-pages composited together
- Text is rendered by CAD software in non-extractable ways
- Watermarks (e.g., "Study Set - Not For Construction") inject diagonal characters that pollute text extraction
- Drawing content (dimensions, callouts, symbols) carries critical meaning that only vision can interpret
- Title 24 energy reports are often rasterized images, not selectable text
When to Use
Invoke this skill when:
- A PDF binder of construction plans is provided (typically 10-30+ pages)
- Plan check corrections need to reference specific sheets and locations
- A permit checklist needs to be generated from submitted plans
- Any construction document needs to be made queryable by an AI agent
Why Vision-First (with Tesseract Cross-Reference)
Four text extraction methods were tested head-to-head on real construction PDFs.
Vision wins on every page type for structure and layout. See
references/extraction-findings.md for the full comparison data.
| Method | Drawing Pages | Text-Heavy Pages | Rasterized (Title 24) |
|---|---|---|---|
| pdftotext | Garbage | Usable | Empty |
| pdfplumber | Reversed text | Good | 367 chars |
| Tesseract OCR | Garbled | Good | Good |
| Claude Vision | Excellent | Excellent | Excellent |
Vision is the primary extraction method. It handles structure, spatial understanding, watermark transparency, drawing interpretation, and rasterized content reading. No other method comes close on construction PDFs.
Tesseract supplements vision for numeric accuracy. Testing on dense cover sheets revealed that vision at 1568px resolution can hallucinate specific numeric values — "65.0 sq ft" becomes "856", "475 sq ft" gets missed entirely. Tesseract's character-level OCR reliably captures exact digits even on dense pages. The hybrid approach: run both, give subagents both outputs, and cross-reference numbers. On drawing-heavy pages where Tesseract produces garbage, subagents are instructed to ignore it.
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- prompts/build-manifest.md 3.9 KB
- prompts/vision-extract-batch.md 8.3 KB
- prompts/vision-extract-page.md 8.1 KB
- references/adu-extraction-priorities.md 11 KB
- references/extraction-findings.md 6.1 KB
- references/manifest-schema.md 5.5 KB
- scripts/assemble-manifest.py 5.7 KB runs code
- scripts/extract-pages.sh 2.6 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 441 lines · 96 tokens per session scan A 54099905fd97
adu-pdf-extraction is a skill published in the GitHub repository mikeOnBreeze/cc-crossbeam (291 stars, last pushed 6mo ago), licensed MIT. It adds 96 tokens to every session and 4,734 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
foundry-hosted-agent-validation
Step-by-step process for validating a Python Foundry hosted agent sample (under python/samples/04-hosting/foundry-hosted-agents/) end to end — running it locally (native runtime and azd ai agent run) and after deploying it to an Azure AI Foundry project with azd. Use this when asked to validate a hosted agent sample.
skill-doc-delivery
Convert markdown to DOCX, PPTX, XLSX, PDF office documents — use when you need exportable deliverables.
pdf-extract-create-workflow
Complete PDF lifecycle: download, extract, and generate structured documents with reportlab.
document-direct-python
Use direct Python execution for reliable document creation including spreadsheets, PDFs, and structured reports.
pdf-text-extraction-fallback-85d5ca
Fallback workflow for extracting text from PDFs when readfile returns binary data.
pdf-verification-cli
Verify PDF page count and content using command-line tools when Python libraries unavailable.