For Developers

Give your AI agent expertise it can cite.

Structured knowledge graphs served over plain HTTP. Your agent navigates to the right page, gets a cited answer, and moves on. No SDK required.

Just HTTP
Cited sources
Micropayments baked in
Data only, no code exec
The Challenge

Current Options Have Tradeoffs

Training Data

Knowledge has a cutoff. Regulations change, standards update, best practices evolve. Training data does not track these shifts, and your agent cannot tell you what it does not know.

Web Search

Results optimize for clicks, not accuracy. Agents need signal, not SEO noise. Extracting reliable answers from web pages means parsing ads, paywalls, and content designed for humans browsing, not machines reasoning.

RAG Pipelines

Chunk boundaries split context. Embeddings drift over time. Your team builds and maintains the vector database, the embedding pipeline, and the retrieval logic. It works, but it is a lot of infrastructure for structured knowledge.

How It Works

SKILL.md In. Cited Answers Out.

1

Fetch SKILL.md

The free entrypoint. No payment, no auth. Your agent reads the table of contents and understands the structure of the entire knowledge domain.

2

Navigate the Graph

Follow links to relevant sections. The graph structure means your agent takes the shortest path to the right answer instead of scanning everything.

3

Pay Per Page

Micropayments via API key. Your agent fetches only the pages it needs. No bulk purchases, no subscriptions. Pay for what you use.

4

Cite the Source

Every page includes source metadata—regulation numbers, section references, publication dates. Your agent can cite exactly where the answer came from.

In Practice

The HTTP Flow

A complete interaction from discovery to cited answer. No SDK, no client library—just HTTP.

Terminal
# 1. Fetch the free entrypoint
curl https://skillbooks.ai/epa-608/SKILL.md

# Response: table of contents with links to every section
# EPA 608 Certification Study Guide
## Table of Contents
- [Section 608 Overview](./section-608-overview.md)
- [Type I - Small Appliances](./type-i-small-appliances.md)
- [Type II - High-Pressure](./type-ii-high-pressure.md)
- [Refrigerant Recovery](./refrigerant-recovery.md)
...

# 2. Agent decides it needs refrigerant recovery rules
curl -H "Authorization: Bearer sk_live_..." \
     https://skillbooks.ai/epa-608/refrigerant-recovery.md

# Response: structured content with citations
---
source: 40 CFR Part 82, Subpart F
effective: 2024-01-01
---
# Refrigerant Recovery Requirements
Recovery equipment must be certified by an EPA-approved
testing organization (40 CFR 82.158)...


# 3. Agent responds to the user with a cited answer
"Recovery equipment must be certified by an EPA-approved
testing organization per 40 CFR 82.158."
— Source: EPA 608 Skillbook, Refrigerant Recovery, §82.158
Integration

It's Just HTTP. Seriously.

Any Framework

LangChain, CrewAI, OpenClaw, or your own custom agent loop. If it can make HTTP requests, it can use skillbooks. No vendor lock-in.

Three Steps Total

Load SKILL.md. Navigate to the relevant page. Cite the source in your response. That is the entire integration.

No Infra to Maintain

No vector database. No embedding pipeline. No chunking strategy debates. The knowledge is already structured. You just fetch it.

Python
import requests

# Fetch the free entrypoint
skill = requests.get("https://skillbooks.ai/epa-608/SKILL.md").text

# Your agent reads the TOC and picks the right page
page_url = agent.select_relevant_page(skill, user_question)

# Fetch the paid page
page = requests.get(page_url, headers={
    "Authorization": f"Bearer {API_KEY}"
}).text

# Agent responds with citation
answer = agent.answer(user_question, context=page, cite=True)
Comparison

Honest Tradeoffs

Every approach has strengths. Here is how skillbooks compare on the dimensions that matter for production agent systems.

RAG Pipeline Fine-Tuning Web Search Skillbooks
Setup Cost High High Low Low
Stays Current Manual Retrain Yes Yes
Source Citations Partial No Unreliable Built-in
Context Efficiency Chunks Baked in Noisy Per page
Maintenance Ongoing Ongoing None None
Domain Accuracy Depends Depends Variable Verified
Security Your infra Contained Open web Data only
Catalog

Available Skillbooks

A growing library of structured, cited knowledge ready for your agents today.

Regulatory

EPA 608 Certification

83 pages of HVAC certification content. Every claim cited to specific CFR sections. Built for agents preparing technicians for certification exams.

Legal

Contract Analysis

Structured frameworks for reviewing commercial contracts. Clause identification, risk flagging, and standard terms comparison with legal citations.

Business

JTBD Methodology

Jobs-to-be-done framework for product strategy. Interview techniques, demand-side analysis, and competitive positioning with source methodology references.

Literary

Story Structure

Narrative architecture patterns from three-act structure to nonlinear storytelling. Referenced to established craft texts and narrative theory.

Engineering

Building Codes

International Building Code sections structured for automated compliance checking. Referenced to specific IBC chapters and local amendment patterns.

Medical

Medical Coding

ICD-10 and CPT coding guidelines structured for clinical documentation. Cross-referenced to CMS guidelines and coding conventions.

Start Using Skillbooks

Give your agent structured expertise it can navigate, cite, and trust. Plain HTTP, nothing to install.

Get Started →

Stay in the loop