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.
git clone --depth 1 https://github.com/hhx465453939/mcp-pubmed-serverWrote 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/rules/hhx465453939/mcp-pubmed-server/fulltext-mode)<a href="https://agentmods.dev/rules/hhx465453939/mcp-pubmed-server/fulltext-mode"><img src="https://agentmods.dev/badge/rules/hhx465453939/mcp-pubmed-server/fulltext-mode.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 5.1 | $0.00000 | $0.02017 |
| Opus 5 | $0.00000 | $0.01009 |
| Sonnet 5 | $0.00000 | $0.00403 |
| Haiku 4.5 | $0.00000 | $0.00202 |
Grade A, and why
fulltext-mode scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
const response = await fetch(downloadUrl, { How it starts
The opening of the file, as written. The whole thing — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.
全文模式开发规范
全文模式架构
环境变量配置
// 全文模式配置
const FULLTEXT_MODE = (process.env.FULLTEXT_MODE || 'disabled').toLowerCase();
const FULLTEXT_ENABLED = FULLTEXT_MODE === 'enabled' || FULLTEXT_MODE === 'auto';
const FULLTEXT_AUTO_DOWNLOAD = FULLTEXT_MODE === 'auto';
模式说明
disabled: 禁用全文功能(默认)enabled: 启用全文检测,手动下载auto: 启用全文检测,自动下载可用的OA论文
全文检测流程
1. 开放获取检测
// 检测OA论文和全文可用性
async detectOpenAccess(article) {
const oaInfo = {
isOpenAccess: false,
sources: [],
downloadUrl: null,
pmcid: null,
doi: article.doi
};
// 1. 检查PMC免费全文
// 2. 检查DOI的Unpaywall
// 3. 检查出版商直接OA
return oaInfo;
}
2. 多源检测策略
// PMC检测
async checkPMCContent(pmid) {
const pmcUrl = `${PMC_BASE_URL}/?term=${pmid}`;
// 检查PMC免费全文可用性
}
// Unpaywall检测
async checkUnpaywall(doi) {
const unpaywallUrl = `${UNPAYWALL_API_URL}/${doi}?email=${email}`;
// 检查Unpaywall数据库
}
// 出版商直接检测
async checkPublisherOA(doi) {
const doiUrl = `https://doi.org/${doi}`;
// 检查出版商网站免费PDF
}
PDF下载和缓存
下载配置
// PDF下载配置
const FULLTEXT_CACHE_DIR = path.join(CACHE_DIR, 'fulltext');
const PDF_CACHE_EXPIRY = 90 * 24 * 60 * 60 * 1000; // 90天过期
const MAX_PDF_SIZE = 50 * 1024 * 1024; // 50MB最大PDF大小
下载流程
// PDF下载方法
async downloadPDF(pmid, downloadUrl, oaInfo) {
const response = await fetch(downloadUrl, {
timeout: 60000, // 60秒超时
headers: {
'User-Agent': 'Mozilla/5.0 (compatible; PubMed-MCP-Server/2.0)',
'Accept': 'application/pdf,text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8'
}
});
// 检查文件大小限制
// 保存PDF文件
// 更新全文索引
}
缓存管理
// 检查PDF是否已缓存
isPDFCached(pmid) {
const pdfPath = path.join(FULLTEXT_CACHE_DIR, `${pmid}.pdf`);
if (fs.existsSync(pdfPath)) {
const stats = fs.statSync(pdfPath);
const age = Date.now() - stats.mtime.getTime();
if (age < PDF_CACHE_EXPIRY) {
return { cached: true, filePath: pdfPath, fileSize: stats.size };
}
}
return { cached: false };
}
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.
- 8d ago First seen · 279 lines · 2,017 tokens per session scan A d1d4471f02f7
fulltext-mode is a cursor rule published in the GitHub repository hhx465453939/mcp-pubmed-server (6 stars, last pushed 6mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 2,017 tokens. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
sumatrapdf-official-toc
Official PDF TOC — a numbering gap means an item was dropped; salvage or reject, never skip the hole.
pro-deck-builder
Create polished HTML slide decks and PDF-ready documents for consulting deliverables. Uses the RRBC design system with warm light mode, dark mode cover pages, Lora/Inter/Roboto Mono typography, and data visualization palette. Trigger on 'deck', 'slides', 'presentation', 'pitch deck', 'keynote', 'report', or 'PDF'.
sigma-reports
Build, validate, retrieve, and safely update Sigma report code representations through /v2/reports/spec. Use for fixed-layout or pixel-perfect reports, invoices, statements, regulatory documents, and PDF delivery. Covers report pages, absolute pixel layout, header/footer panels, common elements, verification…
document-skills
A rule for handling academic papers, Nature-style editing, patents, software documentation, grant materials, presentations, and technical records. It says to choose an installed skill from the /.agents/skills folder and read its instructions first.
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.