DeepPaperNote is an agent skill that reads one research paper and turns its content into a structured Obsidian research note. It is for researchers and learners who want to preserve a paper’s question, methods, evidence, results, and figures in an academic knowledge base. The catalogue entries are its agent skills, instructions, and plugin packaging for use with coding agents.
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 agentmods add instructions/917dhj/deeppapernote/agents-mdgit clone --depth 1 https://github.com/917Dhj/DeepPaperNoteWrote 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/instructions/917dhj/deeppapernote/agents-md)<a href="https://agentmods.dev/instructions/917dhj/deeppapernote/agents-md"><img src="https://agentmods.dev/badge/instructions/917dhj/deeppapernote/agents-md.svg" alt="Measured on agentmods" 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 | $0.00781 | $0.00781 |
| Opus 5 | $0.00391 | $0.00391 |
| Sonnet 5 | $0.00156 | $0.00156 |
| Haiku 4.5 | $0.00078 | $0.00078 |
Grade A, and why
DeepPaperNote AGENTS.md 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 5d 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 — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DeepPaperNote Agent Guide
Repository Purpose
DeepPaperNote is:
- a dual-stack Claude Code and Codex plugin repository
- designed for deep reading of one paper at a time
- intended to support both Claude Code and Codex
- focused on producing a high-quality Obsidian-oriented paper note
DeepPaperNote is not:
- a multi-paper review framework
- a shallow summary generator
Canonical Sources
Use the following source-of-truth hierarchy when working in this repository:
skills/deeppapernote/SKILL.mdis the canonical workflow definitionskills/deeppapernote/scripts/andskills/deeppapernote/references/are part of the reusable coreREADME.mdandREADME.zh-CN.mdare user-facing documentation, not the canonical workflowskills/deeppapernote/agents/openai.yamlis the Codex adapter layer.claude-plugin/plugin.jsonand.codex-plugin/plugin.jsonare host manifests and should stay thin
Adapter layers should stay thin and should not redefine the workflow independently.
Non-Negotiable Product Invariants
Future changes should preserve these product invariants:
- DeepPaperNote handles one paper at a time
- the workflow remains evidence-first
- required stages must not be silently skipped
- weak evidence should fail closed rather than being presented as a full deep read
- figure and table placeholder-first handling remains required
- lint plus final readability review remain required before completion
- Obsidian-first save semantics remain intact
- the paper-local
images/directory remains part of the save protocol - scripts remain deterministic support tools, while paper understanding and final note writing remain model-led
Repository Structure
The repository is layered as follows:
skills/deeppapernote/SKILL.mdis the canonical workflow definition and the skill entrypoint for both Claude Code and other agentsCLAUDE.mdprovides Claude Code project-level guidance and includesAGENTS.mdvia@AGENTS.md.claude-plugin/plugin.jsonidentifies the Claude Code plugin.codex-plugin/plugin.jsonidentifies the Codex pluginskills/deeppapernote/references/stores durable workflow and writing guidanceskills/deeppapernote/scripts/implements the deterministic pipeline and support utilitiesskills/deeppapernote/agents/openai.yamlis the Codex adapterREADME.mdandREADME.zh-CN.mdare user-facing documentation
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.
- 5d ago First seen · 80 lines · 781 tokens per session scan A 51998aecaa0b
DeepPaperNote AGENTS.md is an instructions file published in the GitHub repository 917Dhj/DeepPaperNote (1,031 stars, last pushed 3d ago), licensed MIT. It adds 781 tokens to every session, about $0.0039 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 instructions, from other repositories
llm-wiki-compiler AGENTS.md
AGENTS.md instructions for atomicstrata/llm-wiki-compiler, covering llmwiki, development guidelines, code style & standards, clean code rules and comments and documentation.
pi-llm-wiki AGENTS.md
AGENTS.md instructions for zosmaai/pi-llm-wiki, covering agents.md, project, tech stack, file layout and conventions.
mktero AGENTS.md
AGENTS.md instructions for tenglvjun/mktero, covering agents.md, project overview, commands, repository map and architecture and runtime rules.
hippocampus CLAUDE.md
Instructions for sturlese/hippocampus, covering layout, frontmatter schema, conventions, answering questions from the vault (read order) and contradictions.
onebrain GEMINI.md
Instructions for onebrain-ai/onebrain, covering onebrain — gemini cli, load order, project-level gemini config, tool name mapping and agent instructions.
onebrain AGENTS.md
Instructions for onebrain-ai/onebrain, covering onebrain — codex cli, load order, tool name mapping and agent instructions.