batch-capture

Captures multiple URLs in one job and generates human and agent summary reports.

Updated Apr 23, 2026
One-click install
npx skills add https://github.com/danielrosehill/Claude-Local-Web-Capture-Plugin --skill batch-capture-danielrosehill
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: batch-capture
Source: https://github.com/danielrosehill/Claude-Local-Web-Capture-Plugin/tree/main/skills/batch-capture
Command: npx skills add https://github.com/danielrosehill/Claude-Local-Web-Capture-Plugin --skill batch-capture-danielrosehill

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Capturing and summarizing many web pages one at a time is slow and error-prone, especially for geo-restricted or bot-blocked sites. This Skill processes an entire list of URLs in a single run, saving each article individually and producing a synthesized batch-level report. ## Core Features & Use Cases - Bulk URL capture: Accepts pasted URLs, a text file, or markdown links, deduplicates them, and captures each via an escalation ladder from static HTTP to a real logged-in Chrome session. - Dual report modes: Always writes a human-readable markdown report with per-item cards and a strict-schema JSON report optimized for LLM consumption; optional PDF output via Typst. - Failure transparency: Every input URL appears in the report as captured, thin, or failed, so gaps are never silently dropped. - Use Case: A researcher pastes 15 Israeli news URLs and receives 15 markdown captures plus a thematic summary report and an agent-consumable JSON digest in one batch directory. ## Quick Start Capture all the URLs in this list and give me the individual articles plus a summary report of the whole batch.

Frequently Asked Questions about batch-capture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I capture multiple URLs at once from the command line?▼

Pass the URLs pasted in your message, or point to a text file with one URL per line using the --urls-file flag. The skill deduplicates by normalized URL, captures each page, and writes individual markdown files plus batch reports.

What output formats does batch web capture produce?▼

Each run writes per-article markdown files with frontmatter, a human-readable report.md with per-item cards and a synthesis, and a strict-schema report.agent.json for LLM consumption. An optional report.pdf can be rendered via Typst with the --pdf flag.

Can batch capture handle bot-blocked or paywalled sites?▼

Yes, it uses an escalation ladder: static HTTP fetching first, then stealth headless browsing for bot-blocked pages, and a real logged-in Chrome session via bb-browser for paywalled or session-bound URLs flagged as auth-required.

What happens when some URLs fail during a batch capture?▼

The run continues past individual failures and records each failed URL with its reason in both reports. It only stops early if a systemic issue appears, such as all first-rung fetches failing, which it surfaces for user decision.

Does batch capture send page content to external services for summarization?▼

No, all summarization is done locally by the agent inline. The plugin is localhost-only by design, so no captured content is sent to hosted services.