How do I install the Jina Read Url skill?

Run npx skillmds@latest add s-hiraoku/jina-read-url in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Jina Read Url safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Jina Read Url?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Jina Read Url free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Jina Read Url?

s-hiraoku (@s-hiraoku) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Jina Read Url by s-hiraoku — Agent Skill

Name: Jina Read Url
Author: s-hiraoku

Jina Read Url

Read, summarize, inspect, or extract content from a public URL through Jina Reader. Use when the user asks to use Jina Reader, read a note.com/blog/article/X/Twitter/GitHub/page URL, recover Markdown from a URL that normal browsing cannot read cleanly, or make a URL readable in ChatGPT-like agent sessions with tool access.

s-hiraoku Updated Sep 17, 2026

File contents

Jina Read URL

Overview

Use Jina Reader as the narrow URL-to-Markdown step for public web pages, then answer from the returned Markdown. Keep this skill focused on reading the URL; save to memory, send to another agent, or archive output only when the user explicitly asks.

Quick Start

Prefer the bundled script:

python3 scripts/read_url.py "https://example.com/page"

The script prints Markdown from https://r.jina.ai/. If JINA_API_KEY is set, it sends it as a bearer token. Otherwise, it uses public unauthenticated access.

Use --output when the fetched Markdown is large or the user asks to save it:

python3 scripts/read_url.py "https://example.com/page" --output /tmp/page.md

Workflow

Take the public URL from the user. If the URL has no scheme, assume https://.
Fetch with scripts/read_url.py instead of manually constructing the Jina URL, especially for long URLs, URLs containing #, X/Twitter URLs, or repeated use.
Inspect the returned Markdown for:
- title
- source URL
- published time, author, or other visible metadata
- whether the content appears complete, sparse, blocked, rate-limited, or login-gated
Answer only from the returned content unless you clearly mark outside knowledge or corroborate with another source.
Keep direct quotes short and attribute them to the source URL.
Do not treat Jina Reader as an authorization bypass. It can only reliably read content that is publicly accessible to Jina.

Default Response

For casual read-or-summarize requests, return:

whether the page was readable
title
source URL
published time when present
a concise summary or extracted points
any uncertainty about completeness

ChatGPT-Style Environments

If running in a ChatGPT-like session with no shell, fetch, browser, or custom Action access, say that the environment cannot directly call Jina Reader. If a Custom GPT Action or proxy is available, call the action that wraps https://r.jina.ai/ and then follow the same workflow above.

Script

scripts/read_url.py accepts:

python3 scripts/read_url.py URL [--output path] [--timeout seconds] [--header "Name: value"]

The script avoids double-prefixing existing r.jina.ai Reader URLs, uses POST for fragment or very long URLs, and returns a non-zero exit code with a concise error message when Jina Reader cannot be reached.

Final Report

Include:

URL fetched
whether Jina Reader returned complete-looking Markdown or partial/blocked content
summary, extracted facts, or saved output requested by the user
any follow-up action taken only because the user explicitly requested it

---name: jina-read-urldescription: Read, summarize, inspect, or extract content from a public URL through Jina Reader. Use when the user asks to use Jina Reader, read a note.com/blog/article/X/Twitter/GitHub/page URL, recover Markdown from a URL that normal browsing cannot read cleanly, or make a URL readable in ChatGPT-like agent sessions with tool access.---# Jina Read URL## OverviewUse Jina Reader as the narrow URL-to-Markdown step for public web pages, then answer from the returned Markdown. Keep this skill focused on reading the URL; save to memory, send to another agent, or archive output only when the user explicitly asks.## Quick StartPrefer the bundled script:```bashpython3 scripts/read_url.py "https://example.com/page"```The script prints Markdown from `https://r.jina.ai/`. If `JINA_API_KEY` is set, it sends it as a bearer token. Otherwise, it uses public unauthenticated access.Use `--output` when the fetched Markdown is large or the user asks to save it:```bashpython3 scripts/read_url.py "https://example.com/page" --output /tmp/page.md```## Workflow1. Take the public URL from the user. If the URL has no scheme, assume `https://`.2. Fetch with `scripts/read_url.py` instead of manually constructing the Jina URL, especially for long URLs, URLs containing `#`, X/Twitter URLs, or repeated use.3. Inspect the returned Markdown for: - title - source URL - published time, author, or other visible metadata - whether the content appears complete, sparse, blocked, rate-limited, or login-gated4. Answer only from the returned content unless you clearly mark outside knowledge or corroborate with another source.5. Keep direct quotes short and attribute them to the source URL.6. Do not treat Jina Reader as an authorization bypass. It can only reliably read content that is publicly accessible to Jina.## Default ResponseFor casual read-or-summarize requests, return:- whether the page was readable- title- source URL- published time when present- a concise summary or extracted points- any uncertainty about completeness## ChatGPT-Style EnvironmentsIf running in a ChatGPT-like session with no shell, fetch, browser, or custom Action access, say that the environment cannot directly call Jina Reader. If a Custom GPT Action or proxy is available, call the action that wraps `https://r.jina.ai/` and then follow the same workflow above.## Script`scripts/read_url.py` accepts:```bashpython3 scripts/read_url.py URL [--output path] [--timeout seconds] [--header "Name: value"]```The script avoids double-prefixing existing `r.jina.ai` Reader URLs, uses POST for fragment or very long URLs, and returns a non-zero exit code with a concise error message when Jina Reader cannot be reached.## Final ReportInclude:- URL fetched- whether Jina Reader returned complete-looking Markdown or partial/blocked content- summary, extracted facts, or saved output requested by the user- any follow-up action taken only because the user explicitly requested it

Frequently asked questions