Instances Directory - Project Configurations
This directory contains project-specific configurations that customize Agentic Creator OS behavior.
This directory contains project-specific configurations that customize Agentic Creator OS behavior.
What Are Instances?
Instances define how Agentic Creator OS behaves for different projects:
- Brand voice and tone
- Enabled skills and workflows
- Evaluation settings
- Integration preferences
Directory Structure
instances/
├── CLAUDE.md ← YOU ARE HERE
├── _template/ ← Template for new projects
│ ├── config.json ← Template configuration
│ └── config.ts ← Config utilities
├── frankx/ ← FrankX project
│ └── config.json ← FrankX configuration
└── demo-project/ ← Demo project
└── config.json ← Demo configuration
Instance Configuration
Each instance has a config.json with:
{
"name": "Project Name",
"slug": "project-slug",
"version": "1.0.0",
"brand": {
"name": "Brand Name",
"voice": {
"tone": "professional",
"personality": ["trait1", "trait2"],
"keywords": ["word1", "word2"],
"avoidKeywords": ["word3"]
},
"colors": {
"primary": "#123456",
"secondary": "#654321",
"accent": "#abcdef"
}
},
"directories": {
"workflows": "workflows",
"templates": "templates",
"skills": "skills",
"mcpServers": "mcp-servers"
},
"integrations": {
"opencode": true,
"claudeCode": true,
"gemini": false
},
"skills": {
"enabled": ["content", "marketing"],
"custom": []
},
"workflows": {
"enabled": ["social-media", "content"],
"custom": []
},
"evaluation": {
"enabled": true,
"strictMode": false,
"benchmarkAgainst": "industry-average"
}
}
Available Instances
frankx
Purpose: FrankX brand content
Settings:
- Tone: Provocative, visionary
- Brand: FrankX
- Skills: All enabled
- Evaluation: Strict mode, benchmark against top performers
Use for: Creating FrankX-branded content
demo-project
Purpose: Demo and testing
Settings:
- Tone: Friendly, educational
- Brand: Demo Brand
- Skills: content, marketing only
- Evaluation: Standard mode
Use for: Testing workflows, learning the system
_template
Purpose: Template for new projects
Use for: Creating new project instances
How to use:
cp -r instances/_template instances/my-new-project
# Edit instances/my-new-project/config.json
Creating a New Instance
-
Copy template:
cp -r instances/_template instances/my-project -
Edit config.json:
- Change name, slug
- Set brand voice
- Enable/disable skills
- Configure workflows
-
Use the instance: The system will automatically detect which instance to use based on your project.
Instance Selection
The system selects an instance based on:
- Project directory name - Matches instance slug
- Manual override - Specify in configuration
- Default instance - Falls back to frankx
Brand Voice Configuration
| Setting | Description | Options |
|---|---|---|
tone | Overall voice | professional, casual, authoritative, friendly, provocative |
personality | Traits | innovative, helpful, reliable, cutting-edge, etc. |
keywords | Approved terms | ["AI", "automation", "workflow"] |
avoidKeywords | Terms to avoid | ["spammy", "clickbait"] |
For AI Agents
When working with instances:
- Check the instance - Know which project you're working on
- Respect brand voice - Match the configured tone
- Use enabled skills - Don't use disabled skills
- Follow evaluation rules - Strict mode = higher standards
- Know the keywords - Use approved terminology
Loading Instances
import { getInstanceConfig, listInstances } from './instances/_template/config.ts';
// List all instances
const instances = await listInstances();
// Returns: [{ name, description, path }]
// Load specific instance
const config = await getInstanceConfig('frankx');
// Returns: ProjectConfig object
Related Directories
| Directory | Purpose |
|---|---|
skills/ | Skills customized by instance |
workflows/ | Workflows customized by instance |
templates/ | Templates customized by instance |
mcp-servers/ | Tools customized by instance |
Instances personalize Agentic Creator OS for each project. Know your instance, respect its rules. </claude-mem-context>
Related Documents
Claude AI Git Workflow Integration
When working with git repositories and suggesting commit workflows, always recommend using the `git-ai-commit` CLI tool for creating commit messages.
Code indexing for AI agents: summarization strategies and evaluation systems
**The most effective code indexing systems combine hierarchical LLM-generated summaries with AST structural data and vector embeddings through hybrid retrieval—achieving up to 80% codebase reduction while maintaining high accuracy for AI coding agents.** Leading tools like Cursor, Sourcegraph Cody, and Continue.dev demonstrate that no single retrieval method suffices; production systems require semantic search, keyword matching, and structural queries working together. For evaluation, the field
Missing Business Agents Research — FLUXION 2026
> Deep Research CoVe 2026 | Date: 2026-03-23
write-script
Write a full video script for @SketchySurvival101 following all rules in CLAUDE.md.