skills.rest
letuhao
letuhaoCOMMUNITY
Explore all skills by this author →
GitHub

aif-reference

Create structured knowledge references from URLs and local files for AI agents.

31|2|Updated Mar 21, 2026
Search Tags:#documentation#markdown#knowledge management#rag#web scraping#ai context#reference generator
One-click install
npx skills add https://github.com/letuhao/lore-weave --skill aif-reference-letuhao
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aif-reference
Source: https://github.com/letuhao/lore-weave/tree/main/.cursor/skills/aif-reference
Command: npx skills add https://github.com/letuhao/lore-weave --skill aif-reference-letuhao

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? AI agents often lack access to documentation they were not trained on, forcing users to repeatedly paste context into conversations. This Skill fetches web pages or reads local files and converts them into durable, structured reference documents that any AI workflow can reuse. ## Core Features & Use Cases - Multi-Source Ingestion: Fetch and process web URLs (with sub-page crawling and targeted search) or local files in Markdown, HTML, JSON, YAML, or plain text formats. - Structured Reference Generation: Synthesize sources into a standardized reference format with overview, core concepts, API signatures, usage patterns, configuration tables, best practices, and pitfalls. - Reference Management: List, show, delete, and refresh existing references with an --update mode that re-fetches sources and updates only changed sections. - Use Case: Run the Skill against a library's documentation site to produce a reusable reference file, then have planning and implementation workflows ground their output in that reference instead of relying on model memory. ## Quick Start Ask the AI to create a reference from a documentation URL, for example by providing the docs link and an optional reference name.

Frequently Asked Questions about aif-reference

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

FAQPage Schema
How do I create a knowledge reference from a URL for an AI agent?▼

Provide the documentation URL to the Skill, optionally with a reference name. It fetches the page, follows up to eight critical sub-pages, fills gaps with targeted searches, and saves a structured Markdown reference in the configured references directory.

How do I give AI agents access to documentation they were not trained on?▼

Convert the documentation into a reusable reference file stored in a shared references directory. Other workflows such as planning, implementation, and research skills can then read that file for grounded domain context instead of relying on model memory.

Can I build a reference from local files instead of URLs?▼

Yes. The Skill reads local files in Markdown, HTML, JSON, YAML, or plain text, follows referenced local files up to five levels deep, and synthesizes the content into the same structured reference format used for web sources.

How do I update an existing reference when documentation changes?▼

Run the Skill in update mode with the reference name. It re-fetches the sources listed in the reference header, compares new material against existing content, updates only changed sections, and preserves the original creation date.

What happens when a reference gets too large?▼

References should stay under roughly 1000 lines. When a topic exceeds that limit, the Skill splits it into a directory containing multiple focused reference files linked by a local index file.

Related Skills

Explore More Software Engineering Skills→
agentscope-aiagentscope-ai
34.3k

guidance

Locate and interpret QwenPaw documentation for installation and configuration guidance.

Official
Intermediate
Ard2pArd2p

aif-reference

Fetch, parse, and store structured knowledge references from URLs or local files.

Community
Intermediate
tkmh04tkmh04

create-agents-md

Generate AGENTS.md files with repository and folder-scoped inline rules.

Community
Intermediate
jhl-labsjhl-labs
50

RAG & Vector Search

Retrieve relevant documents and augment LLM prompts with vector search.

Official
Advanced