Extracts and cleans text from JavaScript-heavy Salesforce documentation pages.
npx skills add https://github.com/dcinzona/folios --skill sf-docs
Please help me install this Agent Skill. Skill: sf-docs Source: https://github.com/dcinzona/folios/tree/main/.agents/skills/sf-docs Command: npx skills add https://github.com/dcinzona/folios --skill sf-docs
Many official Salesforce documentation pages are difficult to fetch reliably because they are client-rendered, use nested shadow DOMs, or present shell/placeholder content; this skill provides a deterministic playbook and tools to find and extract the real article content so answers can be grounded in authoritative sources.
Use the sf-docs extractor to fetch a Salesforce doc URL, extract the article title and cleaned body text, and return the exact official URL and source type.
High-intent search queries and answers about installing and using this skill.
Salesforce Help shadow DOM extraction uses browser-rendered fetching with deep shadow DOM traversal to bypass shell pages, returning structured JSON containing the cleaned article body text, title, and source URL.
Developer.salesforce.com serves client-rendered shell pages that standard fetchers cannot execute. Browser-rendered extraction runs the JavaScript and applies heuristics to detect and reject shell content, ensuring only the real article body is captured.
Yes, Playwright can extract content from lightningdesignsystem.com and other official Salesforce doc hosts. The extractor includes child-link awareness to automatically follow and retrieve relevant nested documentation pages.
Ground answers by extracting authoritative text from developer, help, architect, admin, and lightningdesignsystem Salesforce domains. The extractor outputs structured JSON with cleaned text, title, URL, and host type to provide a verified factual basis.
Yes, the extractor includes an optional stealth mode using playwright-stealth. This helps reliably retrieve content from JS-heavy or protected Salesforce documentation pages without triggering browser bot detection blocks.
The extractor returns structured JSON outputs containing the article title, exact source URL, host type, cleaned article body text, and extracted content links to provide a complete and authoritative grounding payload.
ThrookieAutomate stealth web scraping of protected sites using Playwright and stealth plugins.
Kingly-AgencyBuild RAG pipelines that ground LLM responses with retrieved external knowledge.
Harmeet10000Assemble end-to-end RAG pipelines with LangChain loaders, splitters, embeddings, and vector stores.
forcedotcomExtract official Salesforce documentation content from JavaScript-heavy web pages.
