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 jimezsa/opencolab --skill pdf-figure-extractgit clone --depth 1 https://github.com/jimezsa/opencolabWrote 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/jimezsa/opencolab/pdf-figure-extract)<a href="https://agentmods.dev/skills/jimezsa/opencolab/pdf-figure-extract"><img src="https://agentmods.dev/badge/skills/jimezsa/opencolab/pdf-figure-extract/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/jimezsa/opencolab/pdf-figure-extract"><img src="https://agentmods.dev/badge/skills/jimezsa/opencolab/pdf-figure-extract.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 | $0.00040 | $0.02086 |
| Opus 5 | $0.00020 | $0.01043 |
| Sonnet 5 | $0.00008 | $0.00417 |
| Haiku 4.5 | $0.00004 | $0.00209 |
Grade A, and why
pdf-figure-extract 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 9d 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 — 241 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PDF Figure Extract Skill
Use this skill when the user wants a figure, architecture image, pipeline diagram, qualitative result panel, or other visual artifact extracted from a paper that already exists locally under the current project.
Typical use cases:
- "send me the architecture figure from this paper"
- "extract Figure 3 from the local PDF"
- "return the pipeline overview image"
- "find the model diagram from the paper and send it back"
- "pull out the qualitative comparison panel on page 7"
This skill complements pageindex-grounded, not replaces it.
When cached PageIndex artifacts already exist, reuse them to narrow candidate pages before extraction.
When PageIndex is missing or inactive, continue in standalone PyMuPDF mode instead of blocking the run.
Mission
Given a request for a figure in an already-downloaded local PDF:
- Select a bounded local paper set.
- Reuse PageIndex artifacts when available to narrow likely pages, otherwise shortlist pages directly from the PDF.
- Extract or render 1-3 figure candidates with PyMuPDF.
- Inspect those candidates with the agent's multimodal capability when available to confirm which image best matches the user's request.
- Save the exported artifacts and manifest data under the active research run folder, normally
<RUN_ROOT>/figures/. - Return the best figure with paper and page context, plus explicit limitations when confidence is reduced.
Prerequisites
- Local PDFs already exist under an active research run folder, normally
research/<YYYY-MM-DD>-<topic-slug>/pdf/, or under the legacy flatresearch/pdf/layout. python3is installed and available inPATH.- PyMuPDF is installed:
python3 -m pip install PyMuPDF
Optional but recommended:
<RUN_ROOT>/pageindex/manifest.jsonand cached PageIndex trees already exist.- paper summaries exist under
<RUN_ROOT>/pdf/*.md.
If PyMuPDF is missing, only install it when the user explicitly asks for installation or setup work.
What ships with it
1 file 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.
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.
- 9d ago First seen · 241 lines · 40 tokens per session scan A c454a3d9aead
pdf-figure-extract is a skill published in the GitHub repository jimezsa/opencolab (11 stars, last pushed 26d ago), licensed MIT. It adds 40 tokens to every session and 2,086 once invoked, about $0.0002 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
hive.pdf
Read, write, merge, split, rotate, watermark, encrypt, and OCR PDF files using Python (pypdf, pdfplumber, reportlab, pypdfium2) and command-line tools (poppler-utils, qpdf). Use when the user asks to extract text/tables/images from a PDF, create or modify a PDF, combine or split PDFs, OCR a scanned PDF…
recipe-bulk-download-folder
List and download all files from a Google Drive folder.
cli-anything-wps
A command-line tool that controls WPS Office on Windows, including Writer, Calc, and Impress. It can also build presentations from structured JSON data and export them as PowerPoint and PDF files.
solidworks-engineering-drawing
A SolidWorks skill for creating and reviewing engineering drawings, which are manufacturing documents showing views, measurements, holes, parts lists, and title-block information.
mineru-pdf-parser
A PDF conversion tool that uses MinerU to turn complex PDF files into Markdown and structured JSON. It can also extract text, tables, formulas, and images.
nutrient-document-processing
A guide for processing documents through the Nutrient DWS API, an online service for converting, reading, editing, and signing files. It supports PDFs, office documents, web pages, and images.