Agent Handoff Document - ProcessPulse

Project: ProcessPulse (AI-Assisted Writing Process Analyzer)
Last Updated: December 12, 2025
Status: ✅ ANALYZER WORKING + WRITER READY + DOCKER READY
License: Polyform Noncommercial 1.0.0 (Free for education)
GitHub: https://github.com/lafintiger/processpulse

Executive Summary

ProcessPulse is an application for educators to assess student writing by analyzing both the final essay AND the complete AI collaboration history. The core philosophy is 80/20 assessment: 80% of the grade comes from the thinking process, 20% from the final product.

Current State:

Backend (FastAPI) fully functional with assessment pipeline + session storage
Frontend (React) analyzer mode works
Writer interface READY FOR TESTING with all core features
Academic integrity tracking - paste detection, copy tracking, focus monitoring
Export options - DOCX, TXT, HTML, JSON (for assessment)
Perplexica web search - AI-powered research with source citations
Docker deployment - One command to deploy everything (docker-compose up -d)

What's Working Now

Writer Interface (Phase 2 - Complete)

Rich text editor with TipTap (bold, italic, underline, headings, lists, quotes, alignment)
AI chat sidebar with streaming responses
Right-click context menu for Edit with AI, Copy, Cut
Find & Replace (Ctrl+F / Ctrl+H)
Insert links via toolbar
Auto-save to localStorage
Export sessions to backend + local JSON

Academic Integrity Tracking

Paste detection with character/word count (shows % pasted in stats bar)
Copy tracking to detect potential external AI use
Focus tracking - monitors when user leaves the app
Session metrics - typed vs pasted ratio, AI acceptance rate

Backend

FastAPI with SQLite database
Writing session storage (/api/sessions/save, /api/sessions/list, /api/sessions/{id})
Assessment pipeline with RAG
Rubric management

Project Architecture

Directory Structure

Process-Analyzer/
├── app/                          # Python backend
│   ├── __init__.py
│   ├── config.py                 # Settings (Pydantic BaseSettings)
│   ├── api/
│   │   ├── main.py              # FastAPI app + lifespan events
│   │   └── routes/
│   │       ├── health.py        # /health, /api/status endpoints
│   │       ├── models.py        # /api/models (Ollama model list)
│   │       ├── upload.py        # /api/upload/essay, /api/upload/chat-history
│   │       ├── rubric.py        # /api/rubric (get rubric)
│   │       ├── assessment.py    # Assessment endpoints
│   │       └── sessions.py      # NEW: Writing session storage
│   ├── db/
│   │   ├── database.py          # SQLite + SQLAlchemy async setup
│   │   └── models.py            # ORM models (includes WritingSession)
│   └── services/
│       ├── parsing/
│       ├── ollama/
│       ├── rag/
│       ├── rubric/
│       └── assessment/
│
├── frontend/                     # React + Vite + TailwindCSS v4
│   ├── src/
│   │   ├── App.tsx              # Main app - routes between Home/Writer/Analyzer
│   │   ├── index.css            # TailwindCSS + custom styles
│   │   ├── types.ts             # TypeScript interfaces
│   │   ├── components/
│   │   │   ├── Header.tsx
│   │   │   ├── StatusBar.tsx
│   │   │   ├── FileUpload.tsx
│   │   │   ├── AssessmentResults.tsx
│   │   │   ├── ChatViewer.tsx
│   │   │   └── writer/          # Writing interface components
│   │   │       ├── WriterPage.tsx
│   │   │       ├── Editor.tsx        # TipTap editor with all features
│   │   │       ├── ChatSidebar.tsx
│   │   │       ├── InlineEditPopup.tsx
│   │   │       ├── SettingsPanel.tsx
│   │   │       └── index.ts
│   │   ├── lib/
│   │   │   └── ai-providers.ts  # AI provider abstraction
│   │   └── stores/
│   │       └── writer-store.ts  # Zustand state with metrics tracking
│   ├── package.json
│   └── vite.config.ts
│
├── data/
│   └── process_analyzer.db      # SQLite database
│
├── RubricDocs/                   # Assessment rubric documentation
├── Samples/                      # Test files
├── requirements.txt
├── run.py
└── README.md

API Routes

Route	Method	Description
`/health`	GET	Simple health check
`/api/status`	GET	Full system status
`/api/models`	GET	List available Ollama models
`/api/upload/essay`	POST	Upload & parse essay file
`/api/upload/chat-history`	POST	Upload & parse chat history
`/api/rubric`	GET	Get full rubric structure
`/api/sessions/save`	POST	NEW: Save writing session
`/api/sessions/list`	GET	NEW: List all sessions
`/api/sessions/{id}`	GET	NEW: Get session details
`/api/sessions/{id}/export`	POST	NEW: Export for assessment

Event Types (Process Capture)

type EventType =
  | 'session_start'
  | 'session_end'
  | 'text_insert'      // Characters typed
  | 'text_delete'      // Characters deleted
  | 'text_paste'       // Pasted from clipboard (tracks length)
  | 'text_copy'        // Copied to clipboard (potential external AI)
  | 'text_cut'         // Cut to clipboard
  | 'text_select'      // Text selection
  | 'ai_request'       // Asked AI for help
  | 'ai_response'      // AI responded
  | 'ai_accept'        // Accepted AI suggestion
  | 'ai_reject'        // Rejected AI suggestion
  | 'ai_modify'        // Modified AI suggestion
  | 'document_save'
  | 'undo'
  | 'redo'
  | 'focus_lost'       // Window lost focus
  | 'focus_gained'     // Window regained focus

Session Metrics (Academic Integrity)

interface SessionMetrics {
  totalCharactersTyped: number      // Student's original work
  totalCharactersPasted: number     // External content pasted in
  totalCharactersCopied: number     // Content copied out (external AI?)
  aiRequestCount: number            // Times asked AI for help
  aiAcceptCount: number             // AI suggestions accepted
  aiRejectCount: number             // AI suggestions rejected
  focusLostCount: number            // Times switched away from app
  totalFocusLostDuration: number    // Time spent outside app (ms)
}

Database Models

WritingSession (NEW)

class WritingSession(Base):
    id: str                      # UUID
    session_id: str              # Frontend session ID
    document_title: str
    document_content: str        # Final document HTML
    assignment_context: str      # Optional assignment prompt
    word_count: int
    session_start_time: int      # Unix ms
    session_end_time: int        # Unix ms
    events_json: str             # All captured events
    chat_messages_json: str      # AI chat history
    total_events: int
    ai_request_count: int
    ai_accept_count: int
    ai_reject_count: int
    text_insert_count: int
    text_delete_count: int
    ai_provider: str             # ollama, openai, anthropic
    ai_model: str
    status: str                  # active, completed, exported

Running the Application

Backend

cd C:\Users\lafintiger\SynologyDrive\_aiprojects\__Dev\Process-Analyzer
.\venv\Scripts\Activate.ps1
python run.py
# Runs at http://localhost:8000
# API docs at http://localhost:8000/docs

Frontend

cd C:\Users\lafintiger\SynologyDrive\_aiprojects\__Dev\Process-Analyzer\frontend
npm run dev
# Runs at http://localhost:5175

Verify Ollama

curl http://localhost:11434/api/tags

Key Features Implemented

Editor Features

AI Integration

Chat sidebar with streaming
Right-click → Edit with AI
Inline edit popup (Cmd+K)
Multiple providers (Ollama, OpenAI, Claude)
Provider status indicator

Process Tracking

UI/UX

Dark theme
Context menu on right-click
Stats bar (words, time, paste %, AI usage)
Keyboard shortcuts hints

Completed Features (Latest Session)

Feature	Status
Export to DOCX	✅ Done
Export to TXT	✅ Done
Export to HTML	✅ Done
Browser spell check	✅ Done
Keyboard shortcuts help modal	✅ Done
Welcome onboarding modal	✅ Done
Auto-save indicator	✅ Done
Error boundary (crash recovery)	✅ Done
Perplexica Web Search	✅ Done

Remaining Features (Future)

Feature	Priority	Effort
~~Export to PDF~~	✅ Done	-
Focus mode (minimal UI)	Low	Easy
Version history/snapshots	Low	High
Mobile responsiveness	Low	Medium

Configuration

Default Models

Analysis: gpt-oss:latest
Embeddings: bge-m3

Settings Files

Backend: app/config.py
Frontend: frontend/src/stores/writer-store.ts → defaultSettings

⚠️ CRITICAL LESSONS LEARNED - READ BEFORE MAKING CHANGES

1. Model Selection for JSON Output (CRITICAL)

Problem: gpt-oss:latest and some other models do NOT properly support Ollama's format: "json" mode. They return malformed JSON or empty responses, causing all assessment scores to be 0.

Solution: Always use models known to support JSON mode:

✅ qwen3:latest - Works well, fast (4.9GB)
✅ huihui_ai/qwen3-abliterated:8b - Works well, fast (4.7GB)
✅ huihui_ai/qwen3-abliterated:32b - Works well, slower but better quality (18.4GB)
❌ gpt-oss:latest - DOES NOT work with JSON mode

How to test a model's JSON support:

$body = @{model="MODEL_NAME"; prompt="Generate JSON: {score: 75}"; format="json"; stream=$false} | ConvertTo-Json
Invoke-RestMethod -Uri "http://localhost:11434/api/generate" -Method Post -Body $body

2. Ollama Connection Issues

Problem: After a long-running or failed request, Ollama can get stuck and stop responding to new requests.

Symptoms:

API calls to Ollama hang indefinitely
/api/tags endpoint times out
Assessment starts but never progresses

Solution:

Kill Ollama: taskkill /f /im ollama.exe or via Task Manager
Restart Ollama: ollama serve
Verify it's responding: Invoke-RestMethod -Uri "http://localhost:11434/api/tags"

3. Chat Sidebar Scrolling (CSS Flexbox)

Problem: Chat messages scroll off the visible area or don't scroll at all.

Root Cause: Flexbox containers need explicit height constraints for overflow-y: auto to work.

Solution - Apply these CSS rules:

/* Parent container must have fixed/explicit height */
.chat-wrapper {
  height: 100%;      /* or calc(100vh - header) */
  overflow: hidden;  /* Prevent parent from growing */
}

/* Scrollable container needs min-h-0 AND explicit height */
.messages-container {
  flex: 1;
  min-height: 0;       /* CRITICAL for flexbox scrolling */
  overflow-y: auto;
  min-height: 500px;   /* Fallback to ensure visibility */
}

/* Fixed elements (header, input) should not shrink */
.header, .input-area {
  flex-shrink: 0;
}

4. Zustand State Persistence Issues

Problem: State gets stuck in invalid states (e.g., providerStatus: 'checking' forever) due to localStorage persistence.

Solution: When debugging connection issues:

Open DevTools → Application → Local Storage
Clear writer-store key
Refresh the page

5. ProcessPulse Session JSON Format

Problem: The Analyzer didn't recognize session exports from the Writer.

Solution: Added PROCESSPULSE_SESSION format to chat parser that extracts exchanges from both chatMessages and events arrays.

Resolved Issues (Complete List)

Frontend Issues

Black screen after document creation
- Cause: Variable shadowing (document vs global document)
- Fix: Renamed to writerDocument, used window.document
Text selection immediately clearing
- Cause: Auto-updating state on selection change
- Fix: Use right-click context menu instead
Chat sidebar scrolling broken
- Cause: Flexbox containers without proper height constraints
- Fix: Added min-h-0, flex-shrink-0, fixed minHeight on messages container
Resizable panel needed between editor and chat
- Fix: Implemented draggable divider with useState for width, mouse event handlers
Provider status stuck at 'checking'
- Cause: Zustand localStorage persistence with stale state
- Fix: Clear localStorage, also removed excessive console logging

Backend Issues

Unicode emoji errors
- Cause: Windows console encoding
- Fix: Removed all emojis from Python code
Assessment returning all zeros
- Cause: gpt-oss:latest model doesn't support JSON format mode
- Fix: Changed default model to qwen3:latest (or huihui_ai/qwen3-abliterated:32b)
Assessment endpoint was placeholder
- Cause: Original endpoint just returned mock data
- Fix: Implemented full assessment pipeline integration
ProcessPulse session format not recognized
- Cause: Chat parser didn't know about Writer's export format
- Fix: Added PROCESSPULSE_SESSION format with parsing for chatMessages and events
Perplexica CORS errors
- Cause: Browser blocking cross-origin requests to localhost:3000
- Fix: Created backend proxy at /api/perplexica/

Configuration Issues

TailwindCSS v4 config
- Fix: Use @import "tailwindcss" and @tailwindcss/postcss
Vite proxy for API calls
- Added proxy in vite.config.ts for /api to http://localhost:8000

Agent Session Log

Session 1 - December 10, 2025

Agent: Initial PRD Development Agent
Accomplished: Created comprehensive PRD, defined architecture

Session 2 - December 11, 2025 (Morning)

Agent: Development Agent
Accomplished:

Set up complete backend (FastAPI, SQLite, parsers, RAG, assessment)
Set up React frontend with Tailwind v4
Built analyzer UI
Started Phase 2: Writer interface

Session 3 - December 11, 2025 (Afternoon)

Agent: Development Agent
Accomplished:

Fixed black screen bug in Writer
Implemented right-click context menu
Added paste/copy/focus tracking for academic integrity
Added session metrics
Created backend session storage API
Added Find & Replace (Ctrl+F/H)
Added link insertion
Added writing stats bar

Session 4 - December 11, 2025 (Evening)

Agent: Development Agent
Accomplished:

Added DOCX export using docx library
Added TXT export (plain text)
Added HTML export (styled)
Enabled browser spell check in editor
Added keyboard shortcuts help modal (Ctrl+/)
Added welcome onboarding modal for first-time users
Added auto-save indicator with timestamp
Added error boundary component for crash recovery
Added Perplexica Web Search integration
- PerplexicaClient class for local AI-powered search
- SearchPanel component with multiple focus modes (Web, Academic, YouTube, Reddit, Wolfram)
- Results display with expandable sources and citations
- Insert search results directly to chat
- Tracks web_search events for process capture
PROTOTYPE NOW READY FOR STUDENT TESTING

Key Files Added:

frontend/src/lib/export-utils.ts - Export utilities
frontend/src/components/ErrorBoundary.tsx - Error handling
frontend/src/components/writer/SearchPanel.tsx - Perplexica search UI

Key Files Modified:

frontend/src/lib/ai-providers.ts - Added PerplexicaClient class
frontend/src/stores/writer-store.ts - Search state, actions, web_search event type
frontend/src/components/writer/WriterPage.tsx - Export dropdown, shortcuts modal, welcome modal, search button
frontend/src/components/writer/Editor.tsx - Spell check enabled
frontend/src/App.tsx - Wrapped with ErrorBoundary

Session 5 - December 11, 2025 (Late Evening)

Agent: Development Agent
Accomplished:

Fixed Perplexica CORS issue
- Created backend proxy at /api/perplexica/ to bypass browser CORS restrictions
- Updated PerplexicaClient to use backend proxy instead of direct calls
- Perplexica web search now fully functional
Added Polyform Noncommercial 1.0.0 License
- Free for educators, students, educational institutions
- Commercial use requires separate license
Complete Docker Deployment Setup
- Dockerfile - Backend (FastAPI, multi-stage build, ~500MB)
- frontend/Dockerfile - Frontend (React + nginx, ~50MB)
- docker-compose.yml - Full orchestration with:
  - ProcessPulse frontend & backend
  - Ollama with auto model download
  - Perplexica + SearXNG for web search
- env.example - Configuration template
- DEPLOYMENT.md - Detailed deployment guide
- Auto-downloads models on first run (~7.5GB total)
- GPU support option (just uncomment in docker-compose)
- External Ollama support for existing installations

Key Files Added:

Dockerfile - Backend container
frontend/Dockerfile - Frontend container
frontend/nginx.conf - Production nginx config
docker-compose.yml - Full stack orchestration
env.example - Environment configuration
DEPLOYMENT.md - Deployment documentation
LICENSE - Polyform Noncommercial 1.0.0
.dockerignore - Docker build exclusions
app/api/routes/perplexica.py - Perplexica proxy endpoints

Disk Space Requirements (Docker):

Component	Size
Docker images	~2.5 GB
Chat model (llama3.1:8b)	~4.7 GB
Embedding model	~275 MB
Total	~7.5 GB

Session 6 - December 12, 2025

Agent: Development Agent
Accomplished:

Fixed Analyzer Assessment Pipeline
- Assessment endpoint was returning placeholder data - now runs full pipeline
- Added ProcessPulse session format parser (PROCESSPULSE_SESSION)
- Session JSON exports from Writer now parse correctly
Fixed Model JSON Support Issue
- Discovered gpt-oss:latest doesn't support Ollama's format: "json" mode
- Changed default model to qwen3:latest (configurable)
- Documented which models work with JSON mode
Fixed Chat Sidebar Scrolling
- Root cause: Flexbox containers need min-h-0 for scrolling to work
- Added fixed minHeight: 600px to messages area
- Made main page container h-screen overflow-hidden
Added Resizable Chat Panel
- Draggable divider between editor and chat
- Persists width during session
Added Markdown Export
- New exportToMarkdown() function in export-utils.ts
- Converts HTML to clean Markdown format
Fixed AssessmentResults Undefined Errors
- Added null checks for processing_time_seconds, total_score, total_possible
- Added safety for summary and criterion_assessments arrays
Documentation
- Added comprehensive "Critical Lessons Learned" section
- Documented all resolved issues with causes and fixes
- Future agents won't repeat these mistakes

Key Files Modified:

app/api/routes/assessment.py - Full assessment implementation
app/services/parsing/chat_parser.py - ProcessPulse session format
app/config.py - Changed default model to qwen3:latest
frontend/src/App.tsx - Assessment handling, progress display
frontend/src/components/writer/WriterPage.tsx - Resizable panels, markdown export
frontend/src/components/writer/ChatSidebar.tsx - Fixed scrolling
frontend/src/components/AssessmentResults.tsx - Null safety checks
frontend/src/lib/export-utils.ts - Added markdown export

Critical Learnings:

Always test model JSON support before using for assessment
Flexbox scrolling requires min-h-0 on child containers
Ollama can get stuck - restart it if requests hang
Clear localStorage when Zustand state gets corrupted

Session 7 - December 12, 2025

Agent: Development Agent
Accomplished:

Improved Assessment Prompt Strictness
- Updated SYSTEM_PROMPT with explicit red flags for AI over-dependence
- Added copy-paste delegation patterns that trigger automatic INADEQUATE scores
- Added passive consumption patterns that cap scores at DEVELOPING
- Created detailed scoring guidance with examples for each level
- Updated summary prompt to ask "delegation vs collaboration?"
- Updated authenticity prompt with additional delegation checks
- Improved semantic search queries for each criterion
- Result: Test submission scored 59/100 (down from 62/100) - more accurate for copy-paste usage
Added PDF Export for Assessment Reports
- Professional multi-page PDF reports using jsPDF + jspdf-autotable
- Includes: overall score, summary paragraphs, key strengths, areas for growth
- Detailed criterion breakdown with reasoning, evidence citations, feedback
- Authenticity analysis with flags and positive indicators
- Export button added to AssessmentResults component
- Also added JSON data export option
Changed Frontend Port to 5175
- Updated vite.config.ts to use port 5175 (was 5173)
- Avoids conflict with other projects

Session 8 - December 12, 2025 (Current)

Agent: Development Agent
Accomplished:

Made App Remotely Accessible for Student Testing
- Changed backend host from 127.0.0.1 to 0.0.0.0 in app/config.py
- Updated CORS to allow all origins (allow_origins=["*"])
- Frontend runs with --host flag for LAN access
Added Student Information to Document Creation
- New "Student Information" section in New Document modal
- Required: Student Name (must be filled to create document)
- Optional: Student ID
- Student badge appears in editor header showing name/ID
Created Server-Side Submission Storage System
- New /api/submissions/ endpoints in app/api/routes/submissions.py
- Submissions saved to data/submissions/{student_name}/ directory
- Each submission creates two files:
  - {title}_{timestamp}.md - Essay in Markdown with YAML frontmatter
  - {title}_{timestamp}_session.json - Full session data for assessment
- Includes computed stats: AI requests, paste counts, characters typed, etc.
Added "Submit" Button to Writer
- Green "Submit" button in editor header
- Shows submission result modal (success/failure)
- Submits to server automatically - no manual downloads needed!
Created Instructor Submissions Dashboard
- New "Submissions" option on home page (3-column layout)
- Lists all student submissions with metadata
- Search/filter by student name, ID, or document title
- Download individual MD or JSON files
- Delete submissions
- Accessible at home page → "Submissions"

Key Files Added:

app/api/routes/submissions.py - Submission storage API
frontend/src/components/SubmissionsDashboard.tsx - Instructor dashboard

Key Files Modified:

app/config.py - Changed host to 0.0.0.0 for remote access
app/api/main.py - Added submissions router, CORS * origins
frontend/src/stores/writer-store.ts - Added StudentInfo, submitForAssessment()
frontend/src/components/writer/WriterPage.tsx - Student fields, submit button/modal
frontend/src/App.tsx - Added submissions dashboard route

API Endpoints Added:

Route	Method	Description
`/api/submissions/submit`	POST	Submit writing (saves MD + JSON to server)
`/api/submissions/list`	GET	List all submissions (filterable by student)
`/api/submissions/{id}`	GET	Get full submission details
`/api/submissions/{id}/download/{md\|json}`	GET	Download specific file
`/api/submissions/{id}`	DELETE	Delete a submission

Testing the Prototype (Updated)

For Students (Remote Access)

Go to http://{YOUR_IP}:5175 (get IP from instructor)
Click "Writer"
Click "New Document"
Enter your name (required) and optional Student ID
Enter document title and assignment context
Click "Create"
Start writing! Use AI chat sidebar for help
Right-click on selected text for AI editing
Click "Submit" when done - your work is automatically saved to the server!

For Instructors

Start the server with python run.py (backend) and npm run dev -- --host (frontend)
Give students the URL: http://{YOUR_IP}:5175
Submissions are automatically saved to data/submissions/
Go to http://localhost:5175 → "Submissions" to view all student work
Download MD (essay) or JSON (session) files
Use "Analyzer" to assess submissions with the full rubric

Finding Your IP Address

# Windows
Get-NetIPAddress -AddressFamily IPv4 | Where-Object {$_.InterfaceAlias -notlike "*Loopback*"} | Select-Object IPAddress

# Mac/Linux
ip addr show | grep inet | grep -v 127.0.0.1

Next Steps for Future Agent

Connect Writer → Analyzer flow - Button to directly import session for assessment (Use Submissions dashboard instead)
Testing - Unit tests for backend, integration tests for frontend
Deployment - Docker setup, production config ✅ DONE
Mobile responsiveness - Make writer usable on tablets
Test Docker deployment - Verify docker-compose works on fresh machine
Batch assessment - Multiple submissions at once
PDF export - Export essays as PDF (Already have DOCX/MD/HTML export)
Direct analyze from submissions - One-click to load submission into analyzer

Docker Deployment

Quick Start

git clone https://github.com/lafintiger/processpulse.git
cd processpulse
cp env.example .env
docker-compose up -d
# Access at http://localhost

Services Included

Frontend (nginx + React)
Backend (FastAPI)
Ollama (Local AI)
Perplexica (Web Search)
SearXNG (Search Engine)

Configuration

Edit .env to change:

Ports (default: 80, 11434, 3000)
AI models (default: llama3.1:8b, nomic-embed-text)
Debug mode

See DEPLOYMENT.md for full instructions.

This document should be updated whenever significant progress is made.

FilesExpand file tree

Agents.md

Latest commit

History

Agents.md

File metadata and controls

Agent Handoff Document - ProcessPulse

Executive Summary

What's Working Now

Writer Interface (Phase 2 - Complete)

Academic Integrity Tracking

Backend

Project Architecture

Directory Structure

API Routes

Event Types (Process Capture)

Session Metrics (Academic Integrity)

Database Models

WritingSession (NEW)

Running the Application

Backend

Frontend

Verify Ollama

Key Features Implemented

Editor Features

AI Integration

Process Tracking

UI/UX

Completed Features (Latest Session)

Remaining Features (Future)

Configuration

Default Models

Settings Files

⚠️ CRITICAL LESSONS LEARNED - READ BEFORE MAKING CHANGES

1. Model Selection for JSON Output (CRITICAL)

2. Ollama Connection Issues

3. Chat Sidebar Scrolling (CSS Flexbox)

4. Zustand State Persistence Issues

5. ProcessPulse Session JSON Format

Resolved Issues (Complete List)

Frontend Issues

Backend Issues

Configuration Issues

Agent Session Log

Session 1 - December 10, 2025

Session 2 - December 11, 2025 (Morning)

Session 3 - December 11, 2025 (Afternoon)

Session 4 - December 11, 2025 (Evening)

Session 5 - December 11, 2025 (Late Evening)

Session 6 - December 12, 2025

Session 7 - December 12, 2025

Session 8 - December 12, 2025 (Current)

Testing the Prototype (Updated)

For Students (Remote Access)

For Instructors

Finding Your IP Address

Next Steps for Future Agent

Docker Deployment

Quick Start

Services Included

Configuration