Automated summary
Turns an academic PDF paper into a structured Markdown literature note drawn strictly from the paper itself.
Strengths
- Enforces strict traceability: every note element must come from the paper, with no outside knowledge.
- Specifies chunked page reading so long papers can be covered without losing earlier context.
- Supports several PDF lookup modes: direct path, Zotero key, or title/keyword search.
- Automates saving the note to both an Obsidian inbox and the project literature folder.
Limitations
- The Obsidian path and Zotero storage location are hardcoded for one personal setup.
- Only handles PDFs and relies on Claude Code's Read and Glob tools for file access and lookup.
- Note quality depends entirely on accurate PDF reading; no external verification step is included.
Use it when you need to read an academic PDF and quickly produce a structured, source-grounded literature note.
Avoid it when the paper is not a PDF, when the Zotero/Obsidian paths do not match your machine, or when you want a synthesis enriched with outside sources.
Security analysis
SafeThis is a documentation/note-taking workflow: it locates a PDF, reads it in chunks, fills a Markdown template, and saves the result. There is no shell execution, no network access (WebSearch/WebFetch explicitly forbidden), no secret handling, and no destructive or obfuscated operations. The only notable behaviors are unprompted file writes to a hard-coded personal path and broad globbing of a local Zotero library, both low risk and legitimate for the stated purpose.
- •Instructs writing files without asking for permission ('No permission prompts') to a hard-coded absolute path under another user's home directory (/Users/michal/Library/Mobile Documents/...), which is non-portable and may silently create unexpected directories on other machines
- •References tools (Read, Glob) but declares no allowed-tools, so the actual permission surface is left to the agent harness
- •Reads and processes potentially large local PDF libraries (~/Zotero/storage) including globbing across the whole store
Examples
Review the paper at ~/Downloads/smith2023.pdf and create a structured literature note using the paper review template.Take notes on the paper with Zotero key 23SUA9S3 and save the literature note.Read and summarize the paper 'Attention Is All You Need' from my Zotero storage and produce the structured note.name: paper_review description: Thoroughly review an academic paper and create a structured Markdown note using the note_template.md. Use when the user asks to review, read, summarise, or take notes on a paper.
Paper Review
Overview
Produce a thorough, structured literature note for the indicated paper. The note is filled exclusively from the paper itself — do not introduce any external information, prior knowledge, or internet sources. Every claim in the note must be traceable to the paper.
Input
ARGUMENTS may be:
- A file path to a PDF (absolute or relative)
- A Zotero storage key (e.g.
23SUA9S3) — the PDF lives at~/Zotero/storage/<KEY>/*.pdf - A keyword/title to search in
~/Zotero/storage/using Glob
Instructions
Step 1 — Locate the PDF
- If a direct path is given, use it.
- If a Zotero key is given, Glob
~/Zotero/storage/<KEY>/*.pdf. - If a title/keyword is given, Glob
~/Zotero/storage/**/*.pdfand find the closest match. If multiple candidates exist, list them and ask the user to confirm before proceeding.
Step 2 — Read the PDF
- Use the Read tool with
pages:parameter. - First read pages 1–5 to extract metadata (title, authors, year, journal, DOI, abstract).
- Then read the full paper in chunks of up to 20 pages until the entire document is covered.
- Keep a running internal summary after each chunk — do not lose content from earlier chunks.
- For short papers (≤ 20 pages), read in a single call.
Step 3 — Fill the note template
Read note_template.md (in the same folder as this SKILL.md) and fill every field strictly from the paper:
| Field | Source in paper |
|---|---|
| {{title}} | Title page / header |
| {{authors}} | Author list |
| {{year}} | Publication year |
| {{citekey}} | Derive as AuthorYEAR (e.g. Acemoglu2012) |
| {{tags}} | Infer 2–4 topic tags from content (e.g. IP-finance, Basel-III) |
| {{date}} | Today's date in YYYY-MM-DD |
| {{publicationTitle}} | Journal/conference name |
| {{DOI}} | DOI if present, else N/A |
| {{abstractNote}} | Verbatim or lightly condensed abstract |
| {{annotations}} | Key quoted passages with page numbers |
| {{id}} | Zotero key if known, else leave blank |
| {{pdfLink}} | Absolute path to the PDF |
Fill the analytical sections from your reading:
- Key Takeaways — 3–5 bullet points: the most important insights a reader must retain
- Summary — 2–4 paragraph narrative of what the paper does and why
- Methodology — Approach, data sources, analytical methods used
- Main Findings — Numbered list of empirical or theoretical results
- Connections — What the paper supports, contradicts, extends (infer from paper's own literature discussion only)
- Personal Thoughts — Leave as empty bullet (
-) — the user fills this - Questions — 2–3 open questions that arise from reading the paper
- Reading Progress — Tick all checkboxes that were covered during this review
Step 4 — Save the note
Generate the filename as: citekey.md (e.g. Acemoglu2012.md)
Save the completed note to both locations without asking for permission:
-
Obsidian inbox (always):
/Users/michal/Library/Mobile Documents/iCloud~md~obsidian/Documents/Michal/Research/00_Inbox/<citekey>.md -
Project literature folder (always): Detect the current project from the working directory or the path of the PDF. If the PDF is under
projects/<project_name>/, save toprojects/<project_name>/literature/<citekey>.md. If no project context is clear, save toprojects/literature/<citekey>.md. Create theliterature/folder if it does not exist.
Step 5 — Report to user
After saving, report:
- Citekey and title
- Final note path(s)
- Any fields that could not be filled (e.g. DOI not found in paper)
Rules
- Paper only. Do not add information that is not explicitly stated in the paper. If a field cannot be determined from the paper, write
N/Aor leave blank. - No internet. Do not use WebSearch or WebFetch to look up the paper.
- No permission prompts for saving to the Obsidian inbox or the project literature folder.
- For very long papers (> 60 pages), prioritise: abstract, introduction, methodology, results, conclusion — but still attempt all sections.
ARGUMENTS: $ARGUMENTS
From the file to a first result
The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.
View the source and supporting filesThis download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.
1. Download the file
Download SKILL.mdFrom the root of a test project, save the file as .skill-downloads/paper-review-assistant-mrwaked3/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.
Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.
mkdir -p '.skill-downloads/paper-review-assistant-mrwaked3' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/paper-review-assistant-mrwaked3/download' --output '.skill-downloads/paper-review-assistant-mrwaked3/SKILL.md.part' && mv '.skill-downloads/paper-review-assistant-mrwaked3/SKILL.md.part' '.skill-downloads/paper-review-assistant-mrwaked3/SKILL.md'2. Check prerequisites
Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.
No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.
The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.
3. Load the skill
This file does not provide a usable name/description header for this native procedure. It remains available to download and inspect; use the manual method below.
Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.
4. Try a first task
Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:
Read .skill-downloads/paper-review-assistant-mrwaked3/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.
Supabase Docs Authoring
Documentation
Write, edit, and review Supabase documentation content including guides, tutorials, reference docs, and partials. Follow style guides and MDX/TOML validation rules.
Create Architecture Decision Record
Documentation
Create a new Architecture Decision Record with sequential numbering, register it in the AgentDB graph, and link related ADRs.
Plan update with deep research
Documentation
Updates a plan document by launching parallel subagents to analyze a topic across all compiler passes, then consolidates findings into the plan doc.