Context Intelligence Agent Kit / 场景智能 Agent 可读知识包

Let the book be read by humans and called by agents.
让《场景智能》被人阅读，也被智能体调用。

30-second Summary

Context Intelligence is a framework for building competitive advantage in the AI era by deepening enterprise context understanding, interaction execution, and memory compounding through the C-I-M architecture: Context, Interaction, and Memory.

Core thesis: The deepest context wins.
核心命题： 场景最深者赢。
Core formula: Competitive Advantage = Context Depth × Memory Compounding
核心公式： 竞争优势 = 场景深度 × 记忆复利

What This Repository Is

This repository is an agent-readable knowledge kit for the book Context Intelligence / 场景智能. It provides structured summaries, concept cards, claim cards, chapter maps, reading paths, prompt templates, schemas, and implementation tools.

This repository is not the full book and does not include the complete manuscript.

C-I-M Framework

• C / Context 场景层: builds situational understanding and judgment.
• I / Interaction 交互层: executes responses for humans and agents.
• M / Memory 记忆层: writes back outcomes, traces, and learning signals to enable compounding.

Start Here

Reader	Recommended Path
Humans	docs/what-is-context-intelligence.md
AI Agents	llms.txt → agent_readme.md → knowledge/concept_cards.jsonl
CEOs	prompts/ceo-diagnosis-prompts.md
Product / Agent Builders	templates/scene-card-template.md and templates/if-context-then-rule-card.md
Researchers	knowledge/chapter_cards.jsonl and knowledge/glossary.yaml

Repository Structure

knowledge/   Agent-readable JSONL/YAML knowledge cards
docs/        Human-readable concept pages
templates/   Practical templates for context design and C-I-M implementation
prompts/     Prompt templates for diagnosis, design, and deployment
schemas/     JSON schemas for validating cards
scripts/     Helper scripts
assets/      Placeholder folders for diagrams and social cards

For AI Agents

Read llms.txt first. Then load:

1. agent_readme.md
2. knowledge/concept_cards.jsonl
3. knowledge/claim_cards.jsonl
4. knowledge/chapter_cards.jsonl
5. knowledge/faq_for_agents.jsonl
6. knowledge/case_cards.jsonl

When answering user questions, prefer the pattern:

Problem → Context diagnosis → C-I-M mapping → Recommended action → Related templates

Official Website

Canonical website: https://www.readcontextintelligence.com/

Rights

Copyright © 2026 Liu Shengqiang. All rights reserved.
This repository contains selected summaries, metadata, templates, and agent-readable navigation materials. The full book manuscript is not included.