rag-project-remember

RAG Project Remember — Store to Project Knowledge Graph

Store project-specific knowledge — architecture decisions, conventions, dependencies, and requirements — into the project-level LightRAG graph.

Formatting rules

Format each entry as:

[TYPE] Title — YYYY-MM-DD

What: Brief description
Why: The reasoning or constraint behind it
Files: Relevant file paths (if applicable)

Entry types

Type	Use for
ARCHITECTURE	System design choices, component structure, data flow
CONVENTION	Coding standards, naming patterns, file organization rules
DECISION	Technology choices, library selections, approach decisions
DEPENDENCY	External services, APIs, packages, version constraints
CONFIG	Environment setup, build config, deployment settings
REQUIREMENT	Business rules, constraints, acceptance criteria
INSIGHT	Performance findings, gotchas, non-obvious behaviors

How to store

node -e "
const BASE = 'http://YOUR_LIGHTRAG_HOST:YOUR_PROJECT_PORT';
const API_KEY = process.env.LIGHTRAG_API_KEY || 'YOUR_API_KEY';
(async () => {
  const res = await fetch(BASE + '/documents/text', {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
      'X-API-Key': API_KEY
    },
    body: JSON.stringify({
      file_source: '[TYPE] Title — YYYY-MM-DD',
      text: 'Full content here with What/Why/Files structure'
    })
  });
  console.log('Status:', res.status);
})();
"

After inserting

• Confirm to the user: "Stored to project memory: [TYPE] Title"
• Do NOT dump the raw API response

When NOT to use this

Use /rag-remember (personal graph) instead for:

• User preferences — working style, tool preferences, communication style
• Cross-project knowledge — things that apply to all projects
• Personal context — roles, responsibilities, machine config
• People info — contacts, working relationships

This skill is for things specific to the current project that wouldn't be relevant elsewhere.