AIF-BIN Pro CLI

Professional command-line toolkit for semantic document memory.

---

What is AIF-BIN Pro?

AIF-BIN Pro is a command-line tool for converting documents into searchable AI memory files. It generates real vector embeddings for semantic search — find content by meaning, not just keywords.

Key Features:

• Semantic Search — Query your documents by meaning
• Real Embeddings — 384/768-dimension vectors with 5 model options
• Batch Processing — Parallel conversion of entire directories
• Watch Mode — Auto-sync on file changes
• AI Extraction — Use Claude, GPT-4, Gemini, or Ollama for intelligent parsing
• Multiple Exports — JSON, CSV, Markdown, HTML

---

Installation

Clone the Repository

git clone https://github.com/Terronex-dev/aifbin-pro.git
cd aifbin-pro

Using the Wrapper Script

For convenience, a wrapper script is provided. Make it executable first:

chmod +x aifbin-pro

Now you can run the CLI directly:

# Verify installation
./aifbin-pro --help

The examples below will use this wrapper. If you prefer, you can still run python3 cli/aifbin_pro.py or python cli/aifbin_pro.py.

Optional: Virtual Environment (Recommended)

# Create virtual environment
python3 -m venv .venv

# Activate it
source .venv/bin/activate      # Linux/macOS/WSL
.venv\Scripts\activate         # Windows PowerShell

# Install dependencies
pip install -r requirements.txt

# Run
python cli/aifbin_pro.py --help

---

Quick Start

A sample.md file is included for testing.

# Convert the sample file to AIF-BIN with embeddings
./aifbin-pro migrate sample.md -o output/

# Search your memories semantically
./aifbin-pro search "project decisions" -d output/

# View file info
./aifbin-pro info output/sample.aif-bin

# Extract original content
./aifbin-pro extract output/sample.aif-bin

---

Commands

Command	Description
migrate	Convert files to AIF-BIN (v2 binary format) with embeddings
search	Semantic search across your memory files
info	Show file metadata, chunks, and model info
extract	Recover original content from AIF-BIN
export	Export to JSON, CSV, Markdown, or HTML
watch	Auto-sync directory on file changes
diff	Compare two AIF-BIN files
models	List available embedding models
providers	List AI providers for intelligent extraction
ingest	Convert any file with AI extraction
config	Configure AI provider API keys

---

Command Examples

Migrate (Convert Files)

# Single file
./aifbin-pro migrate document.md -o output/

# Entire directory
./aifbin-pro migrate notes/ -o output/

# Recursive with parallel processing
./aifbin-pro migrate notes/ -o output/ -r -p

# Use a different embedding model
./aifbin-pro migrate notes/ -o output/ -m bge-small

Search (Semantic Query)

# Basic search
./aifbin-pro search "budget decisions" -d output/

# Get more results
./aifbin-pro search "project timeline" -d output/ -k 10

# Filter by date
./aifbin-pro search "meeting notes" -d output/ --after 2026-01-01

Watch (Auto-Sync)

# Watch a directory and auto-convert changes
./aifbin-pro watch notes/ -o output/

# Custom check interval (seconds)
./aifbin-pro watch notes/ -o output/ -i 10

Export (Convert to Other Formats)

./aifbin-pro export file.aif-bin -o output.json -f json
./aifbin-pro export file.aif-bin -o output.csv -f csv
./aifbin-pro export file.aif-bin -o output.md -f markdown
./aifbin-pro export file.aif-bin -o output.html -f html

Ingest (AI-Powered Extraction)

# Configure an AI provider first
./aifbin-pro config --provider anthropic --api-key sk-ant-...

# Ingest files with AI extraction
./aifbin-pro ingest documents/ -o output/ -p anthropic

# Use local Ollama (no API key needed)
./aifbin-pro ingest documents/ -o output/ -p ollama

---

Embedding Models

Model	Dimensions	Speed	Description
minilm	384	Fast	Good quality (default)
mpnet	768	Slow	Higher quality
bge-small	384	Fast	Optimized for retrieval
bge-base	768	Slow	Best quality retrieval
e5-small	384	Fast	Microsoft E5

List available models:

./aifbin-pro models

---

AI Providers (for ingest command)

Provider	API Key Required	Best For
anthropic	Yes	Documents, analysis
openai	Yes	General purpose
gemini	Yes	Images, PDFs (vision)
ollama	No (local)	Privacy, offline use
none	No	Basic text extraction

Configure providers:

# Set API key
./aifbin-pro config --provider anthropic --api-key sk-ant-...

# Set default provider
./aifbin-pro config --default anthropic

# View current config
./aifbin-pro config

List available providers:

./aifbin-pro providers

---

File Format

AIF-BIN v2 is a compact binary format with:

• MessagePack encoding (~50% smaller than JSON)
• Fixed-offset headers for fast parsing
• Embedded CRC64 checksums
• Vector embeddings for semantic search
• Original content preservation

[Header: 64 bytes]
  Magic: "AIFBIN\x00\x01"
  Version, Offsets, Size

[Metadata Section]
  MessagePack blob

[Original Raw Section]
  Preserved source file

[Content Chunks]
  Typed chunks with embeddings

[Footer]
  Index + Checksum

---

Troubleshooting

error: externally-managed-environment (Debian/Ubuntu/WSL)

# Option 1: Use --break-system-packages
pip3 install -r requirements.txt --break-system-packages

# Option 2: Use a virtual environment (recommended)
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

ModuleNotFoundError: No module named 'sentence_transformers'

You may need to install the project's dependencies.

pip install -r requirements.txt

Ollama connection refused

Make sure Ollama is running:

# Start Ollama server
ollama serve

# Pull a model if needed
ollama pull llama3.1:8b

---

Related Projects

Project	Description
AIF-BIN Lite	Free CLI (format conversion only, no embeddings)

---

Privacy & Security

• Local-first: All processing happens on your machine
• No telemetry: We don't collect usage data
• BYO API keys: Use your own AI provider credentials
• Open format: AIF-BIN files are portable, no vendor lock-in

---

Legal

• AIF-BIN™ is a trademark of Terronex.dev
• Licensed under the MIT License (see LICENSE)
• See NOTICE for patent and trademark information

---

Links

• Website: terronex.dev/aifbin
• Support: support@terronex.dev
• Twitter: @terronexdev

---

© 2026 Terronex.dev. All rights reserved.