InnerBoard-local

A 100% offline meeting prep assistant that turns your console history into team/manager-ready updates.

InnerBoard-local analyzes your terminal sessions and produces concise, professional talking points for your next standup or 1:1. Everything runs locally on your machine—no data ever leaves your device.

🎯 What It Does

Record Console Sessions: Capture interactive terminal activity with timing
Extract Insights (SRE): Turn raw logs into structured successes, blockers, and resources
Compose Meeting Prep (MAC): Generate team/manager updates and concrete recommendations
100% Local: Uses local LLMs (Ollama); no data leaves your device
Modern CLI: Beautiful terminal interface with progress indicators and rich formatting

🏗️ Architecture

┌─────────────────┐   process    ┌──────────────────┐   generate    ┌─────────────────┐
│  Terminal Logs  │ ────────────▶│  SRE Sessions    │ ────────────▶ │  MAC Meeting    │
│  (raw console)  │              │  (structured     │              │  Prep Output    │
└─────────────────┘              │   insights)      │              └─────────────────┘
         │                       └──────────────────┘                       │
         │                                │                                 │
         ▼                                ▼                                 ▼
┌─────────────────┐              ┌──────────────────┐              ┌─────────────────┐
│ Encrypted Vault │◀─────────────│ Local Ollama     │─────────────▶│ Team Updates    │
│ (SQLite+Fernet) │              │ Model Processing │              │ Recommendations │
└─────────────────┘              │ (localhost:11434)│              │ Manager Reports │
                                 └──────────────────┘              └─────────────────┘

✨ Features

🔒 Industry-Grade Security

PBKDF2 Key Derivation: 100,000 iterations for cryptographically secure key generation
Fernet AES-128 Encryption: Industry-standard encryption for all stored data
Password-Protected Keys: Optional password encryption of master keys with integrity validation
Input Validation: Comprehensive protection against SQL injection, XSS, path traversal attacks
Data Integrity: SHA256 checksums ensure data hasn't been tampered with
Network Isolation: Only allows loopback connections to Ollama (ports 11434)
Secure File Deletion: Overwrites sensitive files before deletion

🧠 AI-Powered Analysis

SRE (Structured Reflection Extraction): Advanced AI extracts structured insights
- Key successes identification with specifics and context
- Blocker identification with impact assessment and resolution hints
- Resource needs assessment and recommendations
- Session summaries with actionable details
MAC (Meeting Advice Composer): Generates professional meeting prep content
- Team updates with progress highlights and next focus areas
- Manager updates with outcomes, risks, and resource needs
- Concrete recommendations with actionable next steps
- Multi-session synthesis for comprehensive reporting

🎨 User-Friendly Low-Touch Experience

Rich Terminal Interface: Beautiful tables, progress bars, and color-coded output
Comprehensive Help: Detailed command help and usage guidance
Error Handling: User-friendly messages with actionable solutions
Progress Indicators: Real-time feedback during AI processing
Multi-format Display: Tables, panels, structured data, and formatted text

⚡ High Performance

Intelligent Caching: TTL-based caching (responses, models, reflections)
Connection Pooling: Efficient Ollama client management (max 5 connections)
SQLite Optimization: WAL mode, foreign keys, indexed queries
GPU Acceleration: Leverages platform capabilities via Ollama
Memory Management: Automatic cache cleanup and size limits
Thread-Safe Operations: All caching and connections are thread-safe

⚙️ Flexible Configuration

Environment Variables: Full configuration via env vars and .env files
Auto-loading: .env files loaded automatically with python-dotenv
Dynamic Model Switching: Easy switching between Ollama models
Validation: Configuration validation with helpful error messages
Runtime Updates: Configuration changes take effect immediately

🚀 Getting Started

Choose the installation method that works best for you:

⚡ Quick Start (Recommended: Native Install)

Run one command per OS to install natively with a virtual environment and local Ollama:

macOS / Linux:

curl -fsSL "https://raw.githubusercontent.com/ramper-labs/InnerBoard-local/main/quickstart.sh" | bash

Windows (PowerShell):

iwr "https://raw.githubusercontent.com/ramper-labs/InnerBoard-local/main/quickstart.ps1" -UseBasicParsing | iex

The script will:

✅ Create/activate a Python virtual environment
✅ Install InnerBoard-local
✅ Ensure Ollama is installed and running locally
✅ Pull the default AI model (gpt-oss:20b)
✅ Initialize your encrypted vault
✅ Run health checks

During initialization you'll be prompted to set a password for your encrypted vault. Optionally, the quickstart can save your password to .env (plaintext) for convenience; you can decline if you prefer not to store it.

That's it! You're ready to start using InnerBoard-local.

📦 Alternative Installation Methods

Option A: Manual Installation (Recommended)

Prerequisites:

Python 3.8+ (download here)
Git (download here)
Ollama (download here)

# 1. Clone the repository
git clone https://github.com/ramper-labs/InnerBoard-local.git
cd InnerBoard-local

# 2. Environment
python3 -m venv .venv
source .venv/bin/activate

# 3. Install InnerBoard
pip install -e .

# 4. Run automated setup
innerboard setup

Option B: Docker Installation (Flaky at the moment)

# 1. Clone the repository
git clone https://github.com/ramper-labs/InnerBoard-local.git
cd InnerBoard-local

# 2. Run Docker setup script
./docker-setup.sh

Or manually:

# Build and start services
docker-compose up -d

# Initialize InnerBoard (run once)
docker-compose exec innerboard innerboard init

🔧 Post-Installation Setup

After installation, verify everything is working:

# If using virtual environment (created automatically on some systems):
source innerboard_venv/bin/activate

# Check system health
innerboard health

# View available commands
innerboard --help

Note: If the quickstart script created a virtual environment (innerboard_venv), you'll need to activate it each time you want to use InnerBoard-local:

# Activate virtual environment
source innerboard_venv/bin/activate

# Use InnerBoard normally
innerboard add "My reflection"
innerboard list

# Deactivate when done
deactivate

Optional: Add this to your ~/.bashrc for convenience:

echo 'alias innerboard="source innerboard_venv/bin/activate && innerboard"' >> ~/.bashrc
source ~/.bashrc

System Requirements:

RAM: 8GB minimum, 16GB recommended
Storage: 20GB free space for AI models
OS: Linux, macOS, or Windows (WSL)

🕹️ Record a Console Session

Use the built-in recorder to capture an interactive shell session. By default, writes are flushed frequently for near-real-time updates; you can disable with --no-flush.

# Start recording (type `exit` to finish)
innerboard record

# Useful options
# --dir PATH      Save session under a custom directory
# --name NAME     Filename (auto-generated if omitted)
# --shell PATH    Shell to launch (defaults to $SHELL or /bin/bash)
# --flush/--no-flush  Control write flushing (default: --flush)

Session artifacts (raw log, timing, segments, SRE JSON) are saved under your app data directory, e.g. ~/.local/share/InnerBoard/sessions/ on Linux/WSL.

🗣️ Generate Meeting Prep

Aggregate all recorded sessions and produce crisp talking points.

# Generate concise meeting prep
innerboard prep

# Include detailed SRE insights (verbose)
innerboard prep --show-sre

✍️ Optional: Save a Private Note

Initialize Vault and Add a Note

# One-time: create your encrypted vault for notes (password prompt)
innerboard init

# Optional: avoid prompts by exporting your password
export INNERBOARD_KEY_PASSWORD="your_secure_password_here"

# Save a short private note alongside your sessions to remind your future self
innerboard add "Investigated auth token validation; planning staging tests next."

🔧 Suggested Workflow

Record meaningful work sessions: innerboard record
Optionally jot private notes to remind yourself: innerboard add "Short note"
Generate prep before standup/1:1: innerboard prep --show-sre

🎯 Next Steps

Before Standup / 1:1

Run innerboard prep for concise talking points
Use --show-sre to include detailed context when needed

Weekly Review

Skim SRE session summaries to spot patterns
Capture follow-ups as private notes with innerboard add "..."

📋 Quick Reference

Installation & Setup

Command	Description	Example
`curl -fsSL https://raw.githubusercontent.com/ramper-labs/InnerBoard-local/main/quickstart.sh \| bash`	One-command installation	Quick start script
`innerboard setup`	Interactive setup wizard	`innerboard setup --docker`
`innerboard health`	Comprehensive health check	`innerboard health --detailed`
`./docker-setup.sh`	Docker deployment setup	Automated Docker setup

Core Commands

Command	Description	Example
`innerboard init`	Initialize encrypted vault	`innerboard init`
`innerboard add "text"`	Add private reflection	`innerboard add "Struggling with auth tokens"`
`innerboard list`	View saved reflections	`innerboard list --limit 10`
`innerboard delete <id>`	Delete specific reflection	`innerboard delete 5 --force`
`innerboard del <id>`	Alias for delete	`innerboard del 5 --force`
`innerboard clear`	Delete ALL reflections	`innerboard clear --force`
`innerboard status`	Vault status and stats	`innerboard status`

Session Recording & Analysis

Command	Description	Example
`innerboard record`	Record terminal session	`innerboard record --name standup`
`innerboard prep`	Generate meeting prep	`innerboard prep --show-sre`
`innerboard models`	List available AI models	`innerboard models`

Help Commands

Command	Description	Example
`innerboard --help`	Show all commands	`innerboard add --help`

🚨 Important Security Notes

Your data stays local: No information leaves your device
Encryption: All reflections are encrypted at rest
Network isolation: Only local Ollama connections allowed during processing
Password protection: Your master key is password-protected
Secure deletion: Sensitive files are overwritten before deletion

Configuration File

# Create your config file
cp env.example .env

Available Models

# See what models are available
innerboard models

# Expected output:
#    Available   
#  Ollama Models
# ┏━━━━━━━━━━━━━┓
# ┃ Model Name  ┃
# ┡━━━━━━━━━━━━━┩
# │ gpt-oss:20b │
# └─────────────┘
#
# Current model: gpt-oss:20b

🛠️ Troubleshooting Guide

Common Issues

Setup Issues

"externally-managed-environment" error:

# The quickstart script automatically handles this by creating a virtual environment
# If you need to do it manually:
python3 -m venv innerboard_venv
source innerboard_venv/bin/activate
pip install -e .

Quick start script fails:

# Try manual installation
git clone https://github.com/ramper-labs/InnerBoard-local.git
cd InnerBoard-local
innerboard setup

Docker setup fails:

# Check Docker status
docker --version
docker-compose --version

# Clean and retry
docker system prune -a
./docker-setup.sh

Health Check Failures

Run detailed health check:

innerboard health --detailed

Common health check issues:

❌ Python Environment: Upgrade to Python 3.8+
❌ Ollama Service: Run ollama serve
❌ AI Model: Run ollama pull gpt-oss:20b
❌ Vault System: Run innerboard init

"Model not available" Error

# Solution: Pull the model
ollama pull gpt-oss:20b

# Check if Ollama is running
ollama list

"Password required" Error

# Set environment variable
export INNERBOARD_KEY_PASSWORD="your_password"

# Or use the --password flag
innerboard status --password your_password

"No such option: --password" Error

# Some commands don't support --password flag
# Use environment variable instead:
INNERBOARD_KEY_PASSWORD="your_password" innerboard list

Permission Errors

# Ensure write permissions in current directory
chmod 755 .

Ollama Not Starting

# Restart Ollama service
# macOS:
brew services restart ollama

# Linux:
sudo systemctl restart ollama

# Or start manually:
ollama serve

Getting Help

# General help
innerboard --help

# Command-specific help
innerboard add --help
innerboard list --help

Backup your vault:

# Your encrypted data is in:
# - vault.db (encrypted reflections)
# - vault.key (encrypted master key)
cp vault.db vault.db.backup
cp vault.key vault.key.backup

Test Coverage Areas

Component	Tests	Status
Security	16 tests	✅ All passing
Caching	17 tests	✅ All passing
Integration	11 tests	✅ All passing
Network Safety	6 tests	✅ All passing
Total	50 tests	✅ All passing

Manual Testing Verified

✅ First-run setup and recording flow
✅ AI analysis with real Ollama models
✅ Security validation (blocks SQL injection, XSS)
✅ Configuration file loading
✅ Error handling and recovery
✅ Data persistence and integrity
✅ Performance optimization
✅ Multi-reflection management

📄 License

Licensed under the Apache License 2.0. See LICENSE for details.

🙏 Acknowledgments

Powered by Ollama for local model serving
Inspired by the need for private, offline meeting preparation
Built with Click for CLI, Rich for UI, Cryptography for security

📊 Example Output Formats

Input Terminal Log:

~/project $ kubectl get pods -n production
NAME                     READY   STATUS    RESTARTS   AGE
auth-service-7d4b8f9c6-x2p9q   1/1     Running   0          2d
~/project $ kubectl describe pod auth-service-7d4b8f9c6-x2p9q -n production
# ... detailed pod information ...
~/project $ git log --oneline -5
a1b2c3d Fix authentication token validation
e4f5g6h Update user service endpoints
# ... more git history ...

SRE Session Output:

[
  {
    "summary": "Investigated authentication service in production, reviewed recent commits for token validation fixes.",
    "key_successes": [
      {
        "desc": "Successfully accessed production Kubernetes cluster",
        "specifics": "kubectl get pods -n production",
        "adjacent_context": "Authentication service pod running stable for 2 days"
      },
      {
        "desc": "Identified recent authentication fixes in git history",
        "specifics": "git log --oneline -5",
        "adjacent_context": "Found commit a1b2c3d fixing token validation"
      }
    ],
    "blockers": [],
    "resources": [
      "kubectl describe pod auth-service-7d4b8f9c6-x2p9q -n production",
      "Git commit a1b2c3d - authentication token validation fix"
    ]
  }
]

MAC Meeting Prep Output:

{
  "team_update": [
    "✅ Successfully accessed production K8s cluster and verified auth service stability",
    "📊 Reviewed recent authentication fixes, found token validation improvements",
    "🎯 Next focus: testing token validation changes in staging environment"
  ],
  "manager_update": [
    "Team has good access to production systems and can debug effectively",
    "Recent authentication fixes appear stable with 2+ days uptime",
    "No current blockers, investigation skills developing well"
  ],
  "recommendations": [
    "Test the authentication token validation fix in staging environment",
    "Document the debugging process for future authentication issues",
    "Schedule follow-up review of authentication service architecture"
  ]
}

Your offline meeting prep companion that stays on your device. ✨

Name		Name	Last commit message	Last commit date
Latest commit History 49 Commits
.github/workflows		.github/workflows
app		app
config_jsons		config_jsons
ft		ft
prompts		prompts
tests		tests
.gitignore		.gitignore
DOCKER_MAKEFILE_GUIDE.md		DOCKER_MAKEFILE_GUIDE.md
Dockerfile		Dockerfile
LICENSE		LICENSE
MANIFEST.in		MANIFEST.in
Makefile		Makefile
README.md		README.md
docker-compose.yml		docker-compose.yml
docker-setup.sh		docker-setup.sh
example.env		example.env
pytest.ini		pytest.ini
quickstart.ps1		quickstart.ps1
quickstart.sh		quickstart.sh
requirements.txt		requirements.txt
setup.py		setup.py

Folders and files

Latest commit

History

Repository files navigation

InnerBoard-local

🎯 What It Does

🏗️ Architecture

✨ Features

🔒 Industry-Grade Security

🧠 AI-Powered Analysis

🎨 User-Friendly Low-Touch Experience

⚡ High Performance

⚙️ Flexible Configuration

🚀 Getting Started

⚡ Quick Start (Recommended: Native Install)

📦 Alternative Installation Methods

Option A: Manual Installation (Recommended)

Option B: Docker Installation (Flaky at the moment)

🔧 Post-Installation Setup

🕹️ Record a Console Session

🗣️ Generate Meeting Prep

✍️ Optional: Save a Private Note

Initialize Vault and Add a Note

🔧 Suggested Workflow

🎯 Next Steps

Before Standup / 1:1

Weekly Review

📋 Quick Reference

Installation & Setup

Core Commands

Session Recording & Analysis

Help Commands

🚨 Important Security Notes

Configuration File

Available Models

🛠️ Troubleshooting Guide

Common Issues

Setup Issues

Health Check Failures

"Model not available" Error

"Password required" Error

"No such option: --password" Error

Permission Errors

Ollama Not Starting

Getting Help

Test Coverage Areas

Manual Testing Verified

📄 License

🙏 Acknowledgments

📊 Example Output Formats

Input Terminal Log:

SRE Session Output:

MAC Meeting Prep Output:

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages