CyborgShell - Quick Reference

Table of Contents

Philosophy

80s BASIC immediate/program mode, but JavaScript
api.print("hello");              // Immediate - runs now
10 api.print("hello");           // Program - stored, run with 'run'

Architecture and Security

Browser-First, Privacy-Focused

Configuration

Configuration

WARNING:

The Backup configuration to Server backs up your API Keys to the current server space on the server. Currently your API keys will be stored in plain text until ZOSCII ITS encoding is implemented. Make sure you wish to do this, and it is within 'your space' or on 'your server'.

All other options within csconfig currently operate in local storage.
csconfig                                # Open configurator
# Options:
# 1. 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

Execution Modes

Immediate Mode (No Line Numbers)

for(let i=0; i<5; i++) api.print(i);    # Executes immediately

Program Mode (With Line Numbers)

10 for(let i=0; i<5; i++) api.print(i); # Stored in editor
20 api.cls();
run                                      # Execute program
list                                     # View program
save mycode                              # Save to cloud

Command Aliases

Action Locomotive Unix Windows Amiga
List cat ls dir dir
Delete era rm del delete
Rename - mv ren rename
Copy - cp copy copy

AI Features

Quick AI Access

!what is the capital of France?
!how do I sort an array in JS?

ChatGPT Command

chatgpt hello                           # Basic query
chatgpt remember the number 42          # Build context
chatgpt what number did I remember?     # Recall context

Named Sessions

chatgpt math: solve 2+2                 # Session "math"
chatgpt story: write about robots       # Session "story"
math: multiply by 5                     # Shorthand
story: continue                         # Shorthand

Session Management

chatgpt sessions                        # List all sessions
chatgpt save math                       # Save session for later
chatgpt load math                       # Load saved session
chatgpt clear                           # Clear default session
chatgpt clear math                      # Clear named session

Multi-AI Services

chatgpt services                        # List: claude, gemini, openai, ollama
chatgpt service openai                  # Switch to OpenAI
chatgpt service claude                  # Switch to Claude
chatgpt service                         # Show current service

Multi-AI Collaboration

✅ Use named sessions when switching services
analysis: openai analyze this algorithm
analysis: claude now document it beautifully
analysis: gemini translate docs to Japanese
chatgpt save analysis-workflow
⚠️ Avoid: switching services in same session (context contamination)

Interactive Mode

chatgpt                                 # Enter interactive mode
# Now omit 'chatgpt' prefix
remember my name is Julian
what's my name?
x                                       # Exit

Local AI (Ollama)

# Configure in csconfig
endpoint: http://localhost:11434/v1/chat/completions
model: llama3.2:latest

# Use for private data, zero cost
chatgpt service ollama
chatgpt analyze this confidential data

RPG Game

rpggpt                                  # Start interactive fiction
# Session saved as character name

Project Setup

Always Start Here

newfile 10              # Creates files 1-10
files
file 1               # Go to file 1 if not there already
Why? Transformers auto-load from file 11+. This gives you 10 working files.

File Management

file N               # Create N files, go to file N
file X               # Switch to file X
files                # List all files with details
filename name.ext    # Name current file
newfile              # Create one new file
list                 # Show current file
list 10-20           # Show lines 10-20
edit 10              # Edit line 10

Linking (Create Pipelines)

# Single input
link TARGET SOURCE transformer args
link 2 1 passthrough
link 2 1 chatgpt openai process this

# Multi-input (CRITICAL: NO SPACES between numbers!)
link 3 1,2 chatgpt openai combine       Correct: 1,2
link 5 2,3,4 csvmerge no-headers        Correct: 2,3,4
link 3 1, 2 chatgpt                     WRONG: spaces fail!

# Drop events (batch processing)
link drop ocr                           # Drag files → auto OCR
⚠️ CRITICAL: Multi-input links use NO SPACES between numbers: 1,2,3 not 1, 2, 3

Project Management

project save name    # Saves as name.prj (automatic extension)
project load name    # Loads name.prj
project name "Title" # Set display name

Dirty Flags and Saving

files                # Shows D flag for processed files
saveall              # Save all dirty files at once
Workflow:
  1. Name files: filename report.txt
  2. Create links and run pipeline
  3. Check: files (see D flags)
  4. Save results: saveall

Transformers

Data Processing

link 2 1 passthrough                    # Direct copy
link 2 1 null                           # Discard
link 3 1,2 filejoin                     # Join (default: newline)
link 3 1,2 filejoin ---SEP---           # Custom separator
link 4 1,2,3 csvmerge                   # Merge CSVs (keep headers)
link 4 1,2,3 csvmerge no-headers        # Skip duplicate headers
link 3 1,2 jsonmerge                    # Merge JSON objects

Analysis

link 2 1 filestats                      # File statistics
link 3 1,2 filediff                     # Line-by-line diff

AI

link 2 1 chatgpt openai translate to Spanish
link 2 1 chatgpt claude analyze code
link 2 1 chatgpt gemini summarize
link 2 1 chatgpt %PROVIDER% use default

# With sessions (context across links)
link 2 1 chatgpt openai session: calc mean
link 3 2 chatgpt openai session: calc standard deviation

link 2 1 translate Japanese             # Translate
link 2 1 blocker                        # Stop if empty

Output

link 2 1 speak accent:en-us
link 2 1 speak accent:ja-jp
link 3 1,2 template                     # Template + JSON data

Common Patterns

Simple Pipeline (A → B → C)

newfile 3
files
file 1
filename input.txt
file 2
filename processed.txt
file 3
filename output.txt

link 2 1 chatgpt openai process this
link 3 2 translate Japanese

project save simple-pipeline

Fan-Out (1 → Many)

newfile 5
files
file 1
filename source.txt
file 2
filename output1.txt
file 3
filename output2.txt
file 4
filename output3.txt
file 5
filename output4.txt

link 2 1 translate Japanese
link 3 1 translate Chinese
link 4 1 translate Spanish
link 5 1 chatgpt openai summarize

project save fan-out

Fan-In (Many → 1)

newfile 5
files
file 1
filename data1.csv
file 2
filename data2.csv
file 3
filename data3.csv
file 4
filename merged.csv
file 5
filename analysis.txt

link 4 1,2,3 csvmerge
link 5 4 chatgpt openai analyze

project save fan-in

Quality Gate

newfile 5
files
file 1
filename content.txt
file 2
filename analysis.txt
file 3
filename gate.txt
file 4
filename approved.txt
file 5
filename final.txt

link 2 1 chatgpt openai rate quality 1-10
link 3 2 chatgpt openai if score >= 8 output YES else NO or STOP
link 4 3 blocker
link 5 4 chatgpt openai polish

project save quality-gate

Multi-AI Workflow

newfile 4
files
file 1
filename code.js
file 2
filename analysis.txt
file 3
filename docs.txt
file 4
filename final.md

link 2 1 chatgpt openai analyze bugs
link 3 1 chatgpt claude document clearly
link 4 2,3 chatgpt gemini create report

project save multi-ai

Workflow Examples

AI Coder (2 minutes)

newfile 3
files
file 1
filename coder.js
file 2
filename coderai.js
file 3
filename backup.js

link 2 1 chatgpt %PROVIDER% process prompts starting with #
link 3 1 passthrough
link 1 2 passthrough

project save aicoder

file 1
edit 10
10 # make a fibonacci function
run

Translation Hub (2 minutes)

newfile 6
files
file 1
filename brief.txt
file 2
filename story.txt
file 3
filename japanese.txt
file 4
filename chinese.txt
file 5
filename spanish.txt
file 6
filename audio.txt

link 2 1 chatgpt openai write short story
link 3 2 translate Japanese
link 4 2 translate Chinese
link 5 2 translate Spanish
link 6 3 speak accent:ja-jp

project save translation-hub

file 1
edit 10
10 # Story about robot learning emotions

Code Reviewer (3 minutes)

file 4
file 1
filename code.js
file 2
filename review.txt
file 3
filename tests.js
file 4
filename report.md

link 2 1 chatgpt openai analyze bugs and improvements
link 3 1 chatgpt openai generate unit tests
link 4 2,3 chatgpt openai comprehensive report

project save code-review

Debugging

File Not Updating?

Multi-Input Not Working?

AI Confused?

chatgpt clear                           # Clear session
newsession: fresh context               # New session

Service Switch Issues?

# Use named sessions when switching
o: openai task
c: claude task
g: gemini task

Tips

  1. Line numbers = stored, no line numbers = immediate
  2. Use named sessions for different tasks/services
  3. Save session templates for reusable workflows
  4. Multi-AI collaboration: each AI for its strengths
  5. Local AI (Ollama) for private data
  6. %PROVIDER% makes projects portable
  7. Multi-input = AND gate: all sources must be dirty
  8. Blocker = quality gate: stops if empty
  9. saveall after processing: one command saves all
  10. Cross-platform: phone, Xbox, PC - same experience

Remember

Quick Start

# Test immediate mode
api.print("Hello from " + navigator.platform);

# Save as program
10 api.print("Hello World");
save hello
run

# AI query
!explain quantum computing simply

# Create pipeline
newfile 3
files
file 1
filename input.txt
file 2
filename output.txt
link 2 1 chatgpt openai improve this
project save pipeline

# Multi-AI workflow
chatgpt analysis: openai analyze this
analysis: claude document it
analysis: gemini translate to Japanese
chatgpt save analysis-flow