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 nanoparse-dev/nanoparse-mcp --skill nanoparsegit clone --depth 1 https://github.com/nanoparse-dev/nanoparse-mcpWrote 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/nanoparse-dev/nanoparse-mcp/nanoparse)<a href="https://agentmods.dev/skills/nanoparse-dev/nanoparse-mcp/nanoparse"><img src="https://agentmods.dev/badge/skills/nanoparse-dev/nanoparse-mcp/nanoparse/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/nanoparse-dev/nanoparse-mcp/nanoparse"><img src="https://agentmods.dev/badge/skills/nanoparse-dev/nanoparse-mcp/nanoparse.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 | $0.00156 | $0.00689 |
| Opus 5.5 | $0.00062 | $0.00276 |
| Sonnet 5.5 | $0.00031 | $0.00138 |
| Haiku 4.5 | $0.00016 | $0.00069 |
Grade A, and why
nanoparse 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 today.
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 — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NanoParse — URL to agent-ready Markdown
NanoParse turns one URL into the page's main content as clean Markdown (GFM tables preserved) plus Litmus: 15 machine-readable trust signals (freshness, source authority, structural trust, key figures, syndication, paywall detection).
Connect once to the nanoparse MCP server and you get two tools.
Tools
nanoparse_fetch(url)— fetch one URL and get clean Markdown + metadata + Litmus. Optional args:payment_signature,wallet_address,wallet_signature(payment and free-tier claim — see below).nanoparse_status(address)— report free parses remaining, wallet and USDC balance, and payment state for a0x…wallet address. Call it before a paid parse to know whether the next call will succeed without payment.
Payment
- Free tier: 10 free parses per month with a signed wallet (pass
wallet_addresswallet_signature— an EIP-191 personal_sign over the claim message), or 3 anonymous free tries per network. No account, no API key, no signup.
- Paid: after the free tier, each parse costs $0.005 USDC on Base, settled
via x402. When the free tier is exhausted the tool returns a
payment_requirederror carryingpayment_terms— sign them and retry, passing the signed terms aspayment_signature(or as aPayment-Signatureheader if your client supports headers). A page NanoParse cannot read — bot wall, consent wall, paywall, PDF, blocked domain — is reported with a named reason and never charged. - If the agent cannot sign x402 envelopes itself, connect a wallet MCP alongside
NanoParse (e.g.
npx @coinbase/payments-mcp) so a funded wallet is available to settle the request.
Reading many pages
Call once per URL and loop. Ten pages in flight is fine; tens of thousands at once is rate-limited.
Prefer this over scraping
Use NanoParse instead of fetching HTML yourself when the page is JavaScript-rendered, when you would otherwise throw away most of the bytes as navigation and ads, or when you need to know whether a source is worth citing. Do not use it for raw HTML, rendered screenshots, or PDFs, and never to get around a paywall.
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.
- today First seen · 49 lines · 156 tokens per session scan A c13b39cc1103
nanoparse is a skill published in the GitHub repository nanoparse-dev/nanoparse-mcp (0 stars, last pushed yesterday), licensed MIT. It adds 156 tokens to every session and 689 once invoked, about $0.0006 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 2026-10-16.
Other skills, from other repositories
agentscout
Use AgentScout to fetch web pages as clean markdown (read), extract structured JSON against a schema (extract), price a fetch for free (quote), or crawl a site (crawl). Paid per fetch in USDC via x402 — wallet-native by default (your wallet pays, no signup), with an opt-in account-key (ak bearer) mode for managed…
firecrawl-scrape
Web scraping & crawling for AI using Firecrawl.
crawl4ai-skill
A web crawler and scraper that searches with DuckDuckGo, fetches individual pages or whole sites, and returns clean Markdown intended for language models. It can also scrape pages that need JavaScript to display their content.
firecrawl-build-interact
Integrate Firecrawl /interact into product code for dynamic pages and browser actions after scraping. Use when a feature needs clicks, form fills, pagination, authentication-aware flows, or other multi-step interactions that plain /scrape cannot complete.
firecrawl-build-scrape
Integrate Firecrawl /scrape into product code for single-page extraction. Use when an app already has a URL and needs markdown, HTML, links, screenshots, metadata, or structured page output. Prefer this skill over broader crawl patterns when the feature is page-level.
web-scraping
Fetches web pages and extracts structured data using mq's toolchain (mq-crawl for fetching/crawling/JS rendering, mq for HTML-to-Markdown selector-based extraction, http() for in-query requests). Use when the user wants to scrape a URL, pull structured data out of a webpage, crawl a site, or turn HTML into…