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 Azhi-ss/academic-figure-skills --skill academic-figure-architecture-extractorgit clone --depth 1 https://github.com/Azhi-ss/academic-figure-skillsWrote 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/azhi-ss/academic-figure-skills/academic-figure-architecture-extractor)<a href="https://agentmods.dev/skills/azhi-ss/academic-figure-skills/academic-figure-architecture-extractor"><img src="https://agentmods.dev/badge/skills/azhi-ss/academic-figure-skills/academic-figure-architecture-extractor/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/azhi-ss/academic-figure-skills/academic-figure-architecture-extractor"><img src="https://agentmods.dev/badge/skills/azhi-ss/academic-figure-skills/academic-figure-architecture-extractor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00048 | $0.01104 |
| Opus 5 | $0.00024 | $0.00552 |
| Sonnet 5 | $0.00010 | $0.00221 |
| Haiku 4.5 | $0.00005 | $0.00110 |
Grade A, and why
academic-figure-architecture-extractor 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 — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Academic Figure Architecture Extractor
Turn supplied figures into grounded structure and reference-style evidence. Keep scientific content separate from transferable visual grammar.
Load only when relevant:
- missing evidence →
references/missing-info-policy.md - local PDF helper →
scripts/extract_pdf_figures.py
Scope and inputs
Accept one or more local images, PDFs, public paper/figure URLs, or an existing render that needs revision. Prefer the original figure image over a screenshot when both exist. PDF extraction finds embedded images or rasterized pages; it is not an architecture detector, and figure classification remains evidence-based judgment.
For an article URL, inspect the page with an available web or browser tool, locate the requested figure and caption, and resolve its direct image or PDF source. Do not classify a paper URL as a repository. If a renderer will need the reference, retain or download it to an absolute local path without overwriting user files.
Obtain candidates
- Local image: inspect it with an available image-viewing tool at original detail.
- Remote figure: open it, preserve its source URL and caption, then inspect the localized image when possible.
- PDF: use a PDF-capable reader for pages and captions. To extract candidates, run the bundled helper from the skill repository when available:
python3 academic-figure-architecture-extractor/scripts/extract_pdf_figures.py \
/absolute/path/to/paper.pdf -o /absolute/path/to/extraction-directory
Useful optional flags are --backend auto|pdfimages|pymupdf, --pages 1-3|all,
--dpi 150, --min-side 300, and --min-pixels 90000. Read the generated
extract-report.json. If extraction tools are unavailable, use directly viewable
pages or ask for an exported figure; never invent a local path.
Record for every candidate: source URL/path, page or figure number, caption, pixel size, local absolute path when available, and keep/drop/uncertain with a reason.
What ships with it
3 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.
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 · 112 lines · 48 tokens per session scan A 5ac7d131df24
academic-figure-architecture-extractor is a skill published in the GitHub repository Azhi-ss/academic-figure-skills (103 stars, last pushed 9d ago), licensed MIT. It adds 48 tokens to every session and 1,104 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
academic-plotting
Generates publication-quality figures for ML papers from research context. Given a paper section or description, extracts system components and relationships to generate architecture diagrams via Gemini. Given experiment results or data, auto-selects chart type and generates data-driven figures via matplotlib/seaborn.…
academic-plotting
Generates publication-quality figures for ML papers from research context. Given a paper section or description, extracts system components and relationships to generate architecture diagrams via Gemini. Given experiment results or data, auto-selects chart type and generates data-driven figures via matplotlib/seaborn.…
ml-paper-writing
Write publication-ready ML/AI/Systems papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM, OSDI, NSDI, ASPLOS, SOSP. Use when drafting papers from research repos, structuring arguments, verifying citations, or preparing camera-ready submissions. Includes LaTeX templates, reviewer guidelines, and citation verification…
ml-paper-writing
Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research repos, structuring arguments, verifying citations, or preparing camera-ready submissions. Includes LaTeX templates, reviewer guidelines, and citation verification workflows.
ml-paper-writing
Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research repos, conducting literature reviews, finding related work, verifying citations, or preparing camera-ready submissions. Includes LaTeX templates, citation verification workflows, and paper…
paper-read
Read, analyze, and extract knowledge from academic papers. Handles PDFs and arXiv links. Produces structured summaries, identifies core contributions, evaluates methodology, and enables cross-paper comparison.