CyborgShell - Configuration Guide

Overview
When Configuration is Needed
Accessing csconfig
Configuration Options
AI Service Setup
Provider Configuration
Backup and Deployment
Advanced Configuration
Security Considerations
Troubleshooting
Complete Examples

Overview

Configuration is OPTIONAL

CyborgShell works immediately for programming without any configuration. You only need to configure csconfig if you want to use AI features.

Registration and Storage

Registration is OPTIONAL

Registration is only required if you want:

Your own private storage space on the server
To share spaces with other users or teams

Without registration: You can still save programs/files in the public space (accessible to everyone on all devices), but you won't have a private space.

⚠️ Security Reality Check:

CyborgShell is MORE secure than downloading and running random .exe files from the internet. Here's why:

✅ Browser Sandbox: Code runs in your browser's security sandbox (can't access your file system, install keyloggers, or modify system files)
✅ No System Access: Cannot install malware, rootkits, or persistent threats on your computer
✅ Limited Scope: Can only access what's in the browser environment (local storage, session data)
✅ Close Tab = Gone: Close the browser tab and any running code immediately stops

That said, be vigilant when running unknown public code:

Malicious code can access data in your browser local storage (including API keys)
Malicious code can delete or modify files in your CyborgShell space
Malicious code can make unauthorized API calls using your keys

Best Practices:

Review code before running it (especially from public spaces)
Use separate API keys for public experimentation vs. production work
Set spending limits on your AI provider accounts
Your space is private unless you explicitly share it

Bottom Line: Running code in CyborgShell is like using Google Drive, GitHub Codespaces, or JSFiddle - far safer than downloading random software. The browser sandbox is your friend!

How Storage Works

Spaces: Think of spaces like folders or drives:

Public Space: Everyone can access (read/write) - great for learning and sharing
Your Personal Space: Private to you (requires registration)
Shared Spaces: Spaces you've been granted access to by others

File Storage:

Programs and files you save are stored on the CyborgShell server
Your space is accessible from any device where you login
Use space command to switch between spaces
Use grant and revoke to share your space with others

See help account and help sharing in CyborgShell for details on registration and space management.

Operating Modes

CyborgShell can be used in several modes:

Anonymous + Public Space: No registration, save to public space (accessible to everyone on all devices)
Registered + Private Space: Your own private space, save programs privately, share with others
With AI (Optional): Add AI capabilities by configuring API keys in browser local storage (works with or without registration)
Self-Hosted: Run your own CyborgShell server for complete control

When Configuration is Needed

No Configuration Required For:

✅ Writing and running JavaScript code
✅ Creating and managing files
✅ Building projects and pipelines (non-AI transformers)
✅ Using passthrough, blocker, filejoin, csvmerge, OCR, etc.
✅ File operations and data processing

Configuration Required For:

⚠️ Using chatgpt command
⚠️ Using ! quick AI queries
⚠️ AI transformers (chatgpt.xfrm, translate.xfrm)
⚠️ Session management with AI services
⚠️ RPG game (rpggpt)

Accessing csconfig

# Enter configuration interface
csconfig

# You'll see a menu:
# 1. Configure autorun program
# 2. Configure handler (ai, rpggpt)
# 3. Configure provider (chatgpt → service mapping)
# 4. View service details
# 5. Add/update service (API keys, endpoints)
# 6. Delete service
# 
# BS. Backup configuration to Server
# DS. Deploy configuration from Server
# BL. Backup configuration to Local Download
# DL. Deploy configuration from Local File
#
# X. Exit

Configuration Options

Option 1: Configure Autorun Program

Set a program to run automatically when CyborgShell starts.

csconfig
1
Enter program name: startup
# Now 'startup' will run on every launch

Use Case: Automatically load your workspace, initialize variables, or set up your environment

Option 2: Configure Handler

Set the default AI handler for the chatgpt command.

csconfig
2
Available handlers:
  - ai (standard AI chat)
  - rpggpt (interactive fiction/RPG)

Enter handler: ai

Option 3: Configure Provider

Map the keyword "chatgpt" to a specific AI service.

csconfig
3
Enter provider service name: openai
# Now 'chatgpt' uses OpenAI by default

# Or use the special keyword:
Enter provider service name: %PROVIDER%
# Uses whichever service you've marked as default

Tip: Use %PROVIDER% in projects to make them portable across different AI services

Option 4: View Service Details

List all configured AI services and their settings (without showing API keys).

csconfig
4

# Output:
Service: openai
  Endpoint: https://api.openai.com/v1/chat/completions
  Model: gpt-4
  Temperature: 0.7
  Top_p: 1
  Frequency_penalty: 0
  Presence_penalty: 0

Service: claude
  Endpoint: https://api.anthropic.com/v1/messages
  Model: claude-3-5-sonnet-20241022
  ...

Option 5: Add/Update Service

Configure API keys, endpoints, and parameters for AI services.

csconfig
5
Enter service name: openai
Enter API key: sk-proj-...
Enter endpoint [default]: 
Enter model [gpt-4]: gpt-4o
Enter temperature [0.7]: 0.8
Enter top_p [1]: 
Enter frequency_penalty [0]: 
Enter presence_penalty [0]: 
Enter max_tokens [4096]: 8192

Option 6: Delete Service

Remove a configured AI service.

csconfig
6
Enter service name to delete: old-service
Service 'old-service' deleted.

AI Service Setup

OpenAI (ChatGPT)

Getting Your API Key

Go to platform.openai.com
Sign in or create account
Navigate to API Keys section
Create new secret key
Copy key (starts with sk-proj-...)

csconfig
5
Service name: openai
API key: sk-proj-YOUR-KEY-HERE
Endpoint: https://api.openai.com/v1/chat/completions
Model: gpt-4o
Temperature: 0.7
Top_p: 1
Frequency_penalty: 0
Presence_penalty: 0
Max_tokens: 4096

Anthropic (Claude)

Getting Your API Key

Go to console.anthropic.com
Sign in or create account
Navigate to API Keys
Create new key
Copy key (starts with sk-ant-...)

csconfig
5
Service name: claude
API key: sk-ant-YOUR-KEY-HERE
Endpoint: https://api.anthropic.com/v1/messages
Model: claude-3-5-sonnet-20241022
Temperature: 0.7
Top_p: 1
Frequency_penalty: 0
Presence_penalty: 0
Max_tokens: 4096

Google (Gemini)

Getting Your API Key

Go to Google AI Studio
Sign in with Google account
Create API key
Copy key

csconfig
5
Service name: gemini
API key: YOUR-GOOGLE-API-KEY
Endpoint: https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent
Model: gemini-pro
Temperature: 0.7
Top_p: 1
Frequency_penalty: 0
Presence_penalty: 0
Max_tokens: 4096

Ollama (Local AI - No API Key Required)

Setup

Install Ollama from ollama.ai
Pull a model: ollama pull llama3.2:latest
Start Ollama service
Configure in csconfig (no API key needed)

csconfig
5
Service name: ollama
API key: [leave blank or type "none"]
Endpoint: http://localhost:11434/v1/chat/completions
Model: llama3.2:latest
Temperature: 0.7
Top_p: 1
Frequency_penalty: 0
Presence_penalty: 0
Max_tokens: 4096

✅ Complete Privacy: Ollama runs locally with zero external API calls

✅ Zero Cost: No API fees, unlimited usage

✅ Air-Gap Capable: Works without internet connection

Provider Configuration

Setting Default Provider

The %PROVIDER% variable allows projects to be portable across different AI services.

# Set default provider
csconfig
3
Enter provider service name: openai

# Now anywhere %PROVIDER% is used, it maps to 'openai'

Using %PROVIDER% in Projects

# In transformer links
link 2 1 chatgpt %PROVIDER% analyze this

# In interactive mode
chatgpt %PROVIDER%: help me debug

# Switching providers changes behavior everywhere
csconfig → 3 → claude
# Now all %PROVIDER% references use Claude

Multi-Service Setup

Configure multiple services and switch between them:

# Configure all three major services
csconfig → 5 → openai → [configure]
csconfig → 5 → claude → [configure]
csconfig → 5 → gemini → [configure]
csconfig → 5 → ollama → [configure]

# Use explicitly in commands
chatgpt service openai
chatgpt analyze with GPT-4

chatgpt service claude
chatgpt document with Claude

chatgpt service gemini
chatgpt translate with Gemini

chatgpt service ollama
chatgpt private analysis with local AI

Backup and Deployment

Backup Configuration to Server (BS)

⚠️ CRITICAL WARNING:

This option backs up your configuration (including API keys) to the CyborgShell server in PLAIN TEXT until ZOSCII ITS encoding is implemented.

Only use if you control the server (self-hosted)
Do NOT use on public/shared servers with sensitive keys
Configuration is stored in "your space" on the server

csconfig
BS
Configuration backed up to server

Deploy Configuration from Server (DS)

Restore configuration from server backup:

csconfig
DS
Configuration deployed from server

Backup Configuration to Local Download (BL)

✅ RECOMMENDED: Safe local backup method

csconfig
BL
# Downloads: cyborgshell-config-backup.json
# Saves to your Downloads folder

Deploy Configuration from Local File (DL)

Restore configuration from local backup file:

csconfig
DL
# Opens file picker
# Select: cyborgshell-config-backup.json
# Configuration restored

Cross-Device Configuration

Scenario: Work on Multiple Devices

On your PC:

csconfig → 5 → configure all services
csconfig → BL → download backup

On your phone:

# Transfer cyborgshell-config-backup.json to phone
csconfig → DL → select file → deploy

Result: Identical configuration on both devices

Advanced Configuration

Temperature Settings

Controls randomness/creativity (0.0 to 2.0):

0.0-0.3: Deterministic, factual, consistent
0.4-0.7: Balanced (default)
0.8-1.2: Creative, varied
1.3-2.0: Highly random, experimental

# Factual coding assistant
Temperature: 0.2

# Creative writing
Temperature: 1.0

# Experimental exploration
Temperature: 1.5

Max Tokens

Maximum response length:

512-1024: Short responses
2048-4096: Standard (default)
8192-16384: Long-form content
32768+: Very long documents (if supported)

Custom Endpoints

Use custom or proxy endpoints:

# Self-hosted Ollama
Endpoint: http://192.168.1.100:11434/v1/chat/completions

# OpenAI-compatible API
Endpoint: https://your-proxy.com/v1/chat/completions

# Azure OpenAI
Endpoint: https://YOUR-RESOURCE.openai.azure.com/openai/deployments/YOUR-DEPLOYMENT/chat/completions?api-version=2024-02-15-preview

Service-Specific Models

Service	Model Examples	Use Case
OpenAI	`gpt-4o` `gpt-4-turbo` `gpt-3.5-turbo`	Best overall Fast, capable Cost-effective
Claude	`claude-3-5-sonnet-20241022` `claude-3-opus-20240229` `claude-3-haiku-20240307`	Balanced Most capable Fastest
Gemini	`gemini-pro` `gemini-ultra`	Standard Advanced
Ollama	`llama3.2:latest` `mistral:latest` `codellama:latest`	General purpose Fast, efficient Code-focused

Security Considerations

API Key Storage

✅ Your API keys are stored in browser local storage:

Never transmitted to CyborgShell servers*
Never stored in CyborgShell databases
Persist across sessions on same browser
Can be backed up/deployed via csconfig

* unless you use the optional CORS proxy server

Public/Shared Computers

If using CyborgShell on a public computer:

Use private/incognito mode, OR
Clear configuration after use:
```
csconfig → 6 → delete all services
```
Browser local storage is tied to the device

Configuration Backup Security

Method	Security Level	Use Case
BL (Local Download)	✅ Secure	Personal backups, cross-device sync
DL (Local Deploy)	✅ Secure	Restore from personal backup
BS (Server Backup)	⚠️ Use with caution	Only on self-hosted servers
DS (Server Deploy)	⚠️ Use with caution	Only on self-hosted servers

Best Practices

Use separate API keys for dev/prod
Set spending limits on API provider accounts
Backup configuration locally (BL) before major changes
Use Ollama for sensitive/confidential data
Never commit configuration files to version control
Rotate API keys periodically

Troubleshooting

AI Commands Not Working

# Check if services configured
csconfig → 4

# If no services shown, configure at least one
csconfig → 5 → [configure service]

# Set default provider
csconfig → 3 → openai

Invalid API Key Error

# Update service with correct key
csconfig → 5 → [service name] → [paste correct key]

# Verify at provider website:
# OpenAI: platform.openai.com
# Claude: console.anthropic.com
# Gemini: makersuite.google.com

Configuration Not Persisting

# Check browser local storage enabled
# Settings → Privacy → Allow local storage

# Backup configuration
csconfig → BL

# If still issues, try:
# 1. Clear browser cache
# 2. Redeploy configuration: csconfig → DL

Multiple Devices Out of Sync

# On primary device
csconfig → BL → download backup

# Transfer file to other device

# On other device
csconfig → DL → select backup file

Complete Examples

Example 1: First-Time Setup (OpenAI)

# Step 1: Get API key from platform.openai.com
# Step 2: Configure service
csconfig
5
Service name: openai
API key: sk-proj-abc123...
Endpoint: [press Enter for default]
Model: gpt-4o
Temperature: 0.7
Top_p: [press Enter]
Frequency_penalty: [press Enter]
Presence_penalty: [press Enter]
Max_tokens: 4096
X

# Step 3: Set as default provider
csconfig
3
Provider: openai
X

# Step 4: Test
chatgpt hello world
!what is the capital of France?

Example 2: Multi-Service Setup

# Configure OpenAI
csconfig → 5 → openai → [configure]

# Configure Claude
csconfig → 5 → claude → [configure]

# Configure Gemini
csconfig → 5 → gemini → [configure]

# Configure Ollama (local)
csconfig → 5 → ollama → [configure]

# Set default
csconfig → 3 → openai

# Backup everything
csconfig → BL

# Now use different services
chatgpt service openai
!quick question with GPT

chatgpt service claude
!another question with Claude

chatgpt service ollama
!private question with local AI

Example 3: Team Configuration

# Tech lead creates master config
csconfig → 5 → openai → [team API key]
csconfig → 5 → claude → [team API key]
csconfig → 3 → openai
csconfig → BL → save as "team-config.json"

# Share team-config.json with team

# Each team member deploys
csconfig → DL → select "team-config.json"

# Everyone has identical setup

Example 4: Privacy-First Setup (Ollama Only)

# Install Ollama locally
# Visit ollama.ai, download, install

# Pull model
ollama pull llama3.2:latest

# Configure in CyborgShell
csconfig
5
Service name: ollama
API key: [leave blank]
Endpoint: http://localhost:11434/v1/chat/completions
Model: llama3.2:latest
Temperature: 0.7
Top_p: 1
Frequency_penalty: 0
Presence_penalty: 0
Max_tokens: 4096
X

# Set as default
csconfig
3
Provider: ollama
X

# Test - zero external calls
chatgpt analyze this confidential data
# All processing happens locally

# Perfect for:
# - Healthcare (PHI data)
# - Legal (privileged communications)
# - Financial (confidential reports)
# - Classified work (air-gapped environments)

Example 5: Development vs Production

# Development configuration
csconfig → 5 → openai-dev → [dev API key]
csconfig → 5 → ollama → [local setup]
csconfig → 3 → ollama
csconfig → BL → save as "dev-config.json"

# Production configuration
csconfig → 5 → openai-prod → [prod API key]
csconfig → 5 → claude-prod → [prod API key]
csconfig → 3 → openai-prod
csconfig → BL → save as "prod-config.json"

# Switch between environments
# Development:
csconfig → DL → dev-config.json

# Production:
csconfig → DL → prod-config.json

Example 6: Cost-Optimized Setup

# Free tier / low cost
csconfig → 5 → ollama → [local, free]
csconfig → 5 → openai → [gpt-3.5-turbo, cheap]

# Use Ollama for bulk work
chatgpt service ollama
# Process 1000 documents - zero cost

# Use OpenAI for critical tasks
chatgpt service openai
# Final review with GPT-4

Quick Start Checklist

For Programming Only (No AI)

✅ No configuration needed - start coding immediately!

            For AI Features
            ☐ Get API key from provider (OpenAI/Claude/Gemini) or install Ollama
☐ Run csconfig
☐ Select option 5 (Add/update service)
☐ Enter service details
☐ Select option 3 (Configure provider)
☐ Set default service
☐ Exit (X)
☐ Test: chatgpt hello world
☐ Backup: csconfig → BL

        

Summary

Key Points:

Configuration is optional - only needed for AI features
API keys stored in browser local storage (not on servers)
Use BL/DL for safe backup/restore
Use %PROVIDER% for portable projects
Use Ollama for complete privacy (zero external calls)
Multiple services can coexist - switch as needed

CyborgShell - Configuration Guide

Table of Contents

Overview

Registration and Storage

How Storage Works

Operating Modes

When Configuration is Needed

No Configuration Required For:

Configuration Required For:

Accessing csconfig

Configuration Options

Option 1: Configure Autorun Program

Option 2: Configure Handler

Option 3: Configure Provider

Option 4: View Service Details

Option 5: Add/Update Service

Option 6: Delete Service

AI Service Setup

OpenAI (ChatGPT)

Getting Your API Key

Anthropic (Claude)

Getting Your API Key

Google (Gemini)

Getting Your API Key

Ollama (Local AI - No API Key Required)

Setup

Provider Configuration

Setting Default Provider

Using %PROVIDER% in Projects

Multi-Service Setup

Backup and Deployment

Backup Configuration to Server (BS)

Deploy Configuration from Server (DS)

Backup Configuration to Local Download (BL)

Deploy Configuration from Local File (DL)

Cross-Device Configuration

Scenario: Work on Multiple Devices

Advanced Configuration

Temperature Settings

Max Tokens

Custom Endpoints

Service-Specific Models

Security Considerations

API Key Storage

Public/Shared Computers

Configuration Backup Security

Best Practices

Troubleshooting

AI Commands Not Working

Invalid API Key Error

Configuration Not Persisting

Multiple Devices Out of Sync

Complete Examples

Example 1: First-Time Setup (OpenAI)

Example 2: Multi-Service Setup

Example 3: Team Configuration

Example 4: Privacy-First Setup (Ollama Only)

Example 5: Development vs Production

Example 6: Cost-Optimized Setup

Quick Start Checklist

For Programming Only (No AI)

For AI Features

Summary