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 AOROM/paperreading --skill papers-reading-skillgit clone --depth 1 https://github.com/AOROM/paperreadingWrote 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/aorom/paperreading/papers-reading-skill)<a href="https://agentmods.dev/skills/aorom/paperreading/papers-reading-skill"><img src="https://agentmods.dev/badge/skills/aorom/paperreading/papers-reading-skill/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/aorom/paperreading/papers-reading-skill"><img src="https://agentmods.dev/badge/skills/aorom/paperreading/papers-reading-skill.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 | $0.00115 | $0.01366 |
| Opus 5.5 | $0.00046 | $0.00546 |
| Sonnet 5.5 | $0.00023 | $0.00273 |
| Haiku 4.5 | $0.00012 | $0.00137 |
Grade A, and why
papers-reading-skill 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 1mo 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 — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PaperReading Core Adapter
Workflow
- Establish the supplied-source boundary. Identify which title, author, journal, date, DOI, page, section, table, figure, equation, appendix, and full-text surfaces are actually available.
- For UTF-8 text, Markdown, or a text-based PDF, run
paperreading ingest <source> --output <document.json>. Use the resulting source, page, section, block, character-range, and hash metadata; do not invent locators. A scanned PDF requires a separately authorized OCR workflow. - Read the abstract and introduction, theory or hypotheses, data and variables, research design, baseline results, mechanisms, heterogeneity, robustness or endogeneity checks, conclusion, and relevant appendix material.
- Build a v0.3
PaperPackageusingreferences/paper-record.md. When provider output is available, preserve it throughextract,review, andfinalize; never bypass unresolved Candidate or Conflict state. KeepGroundedPaperRecord,EvidenceSpan, andResearchAnalysisconceptually and structurally separate. - For an existing v0.2 record, run
paperreading migrate <record.json> --output <package.json>rather than rewriting it manually. Preserve the resulting migration notes and unverified state. - Run
paperreading validate <package.json>. When a matchingPaperDocumentexists, also runpaperreading verify <package.json> --document <document.json> --strict --output <verified.json>. - Resolve every validation error before export. Keep partial or failed verification visible; never upgrade a package to verified by assertion.
- Produce only the user-selected output. Load
references/reading-fields.mdwhen reviewing the legacy 13-field projection or workbook style.
Evidence protocol
- Record a source claim only when the paper explicitly states it and bind it to at least one evidence ID.
- Record a finding only when supplied text, table, figure, equation, or appendix material supports it.
- Keep researcher or AI-assisted assessment outside the source-grounded record.
- Preserve null results, caveats, sample limits, identification limits, conflicts, and unresolved fields.
- Never fabricate a quotation, page, section, block, table, column, figure, equation, appendix, variable, model, test, ranking, or causal claim.
- Treat traceability as locator specificity, not truth, study quality, causal validity, or external validity.
- Treat source-content verification as a resolution check against the supplied document, not proof that the paper's methods or conclusions are correct.
- Do not translate one journal-ranking system into another. Record only independently verified labels with their source and version boundary.
What ships with it
5 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.
- 1mo ago First seen · 111 lines · 115 tokens per session scan A 3d044414498c
papers-reading-skill is a skill published in the GitHub repository AOROM/paperreading (11 stars, last pushed 12d ago), licensed MIT. It adds 115 tokens to every session and 1,366 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
networkx
Creates, analyzes, and visualizes complex networks and graphs in Python with NetworkX. Use when working with network/graph data structures, computing graph algorithms (shortest paths, centrality, clustering), detecting communities, generating synthetic networks (random, scale-free, small-world), reading/writing graph…
geopandas
Guidance and local audit tools for Python workflows that directly use GeoPandas GeoSeries, GeoDataFrame, spatial operations, or vector-data I/O.
geopandas-spatial
Geospatial and climate data analysis with geopandas and xarray. Use when: (1) geographic vector data analysis, (2) climate and weather NetCDF data, (3) spatial joins and overlay operations, (4) map visualization, (5) CRS transformations and area calculations. NOT for: satellite imagery ML (use rasterio/torchgeo)…
pymatgen-materials
Materials science computation with pymatgen. Use when: (1) crystal structure creation and manipulation, (2) phase diagram construction, (3) electronic structure analysis, (4) symmetry and space group operations, (5) VASP input/output parsing. NOT for: molecular chemistry (use rdkit-chemistry), protein structure (use…
rdkit-chemistry
Molecular chemistry operations via RDKit. Use when: user asks about molecular structures, SMILES, chemical properties, or fingerprints. NOT for: reaction databases or wet lab protocols.
formulate-hypotheses
Generate testable hypotheses from theory, empirical regularity, anomaly, or explicit explanatory competition. Deductive, inductive, abductive, and competing-hypothesis generation are modes; discriminating predictions and comparison are activated in competing mode.