Neura Market
SmolSignal – signal copilot for Flipper Zero files logo

SmolSignal – signal copilot for Flipper Zero files

Free
FreeFree tier
Type
Open Source
Company
SmolNero
Visit Website GitHub (2 stars)

About SmolSignal – signal copilot for Flipper Zero files

SmolSignal is an AI signal copilot designed for Flipper Zero users. It reads Flipper capture files (such as .ir, .sub, .nfc, .rfid, .ibtn, and .txt) and provides plain-English explanations, safety risk classification, and optional integration with local or cloud AI models (Ollama, GPT/OpenAI, DeepSeek, Qwen/DashScope, llama.cpp, vLLM, or any OpenAI-compatible endpoint). The tool fingerprints captures with a Shazam-style identity card, confidence score, and warnings, and maintains a local protocol/device category database to distinguish safe consumer IR, passive sensors, access credentials, vehicle/security signals, GPIolabs, and unknown RF. It features a deterministic safety gate that classifies workflows as safe, explain-only, blocked, or unknown, and can generate Flipper-compatible .ir files for safe consumer IR remotes. Authorized Lab Mode allows controlled use for owned, simulated, or isolated lab work without removing hard safety boundaries—hard-blocked categories (e.g., car keys, access credentials, bypass/replay workflows) remain blocked even in lab mode.

Key Features

Reads common Flipper text captures: .ir, .sub, .nfc, .rfid, .ibtn, .txt
Detects signal domain: infrared, Sub-GHz, NFC, RFID, iButton, GPIO, BLE, or unknown
Offline deterministic readout in beginner-friendly language
Optional AI model integration via Ollama, GPT/OpenAI, DeepSeek, Qwen/DashScope, llama.cpp, vLLM, or any OpenAI-compatible endpoint
Shazam-style fingerprint with identity card, confidence score, signature, evidence, and warnings
Local protocol/device category database for safety classification
Passive sensor mode for safe RF documentation without replay/transmit
Local vector/RAG search over built-in safety and protocol notes
Photo context, Web Serial connection status, and community profile matching
Workflow classification: safe, explain-only, blocked, or unknown

Pros & Cons

Pros
  • Open source and free to use
  • Deterministic safety gate prevents accidental misuse
  • Supports multiple Flipper file formats and signal domains
  • Can use local or cloud AI models for richer explanations
  • Authorized Lab Mode allows flexible experimentation within safe boundaries
  • No AI model required for basic offline functionality
Cons
  • Requires Node.js 20+ and npm 10+ to run locally
  • Limited to Flipper Zero capture files; no support for other SDR or hardware formats
  • Hard-blocked categories may restrict legitimate research without proper scope notes
  • Dependent on external AI model endpoints for enhanced explanations

Best For

Analyzing and understanding Flipper Zero capture filesLearning about signal types and safety risks from captured dataGenerating safe consumer IR remote control filesDocumenting and annotating captures for security research or educationUsing AI to enrich explanations while keeping safety deterministic

FAQ

What file formats does SmolSignal support?
SmolSignal reads common Flipper-style text captures: .ir, .sub, .nfc, .rfid, .ibtn, and .txt files.
Can SmolSignal transmit or replay signals?
No. SmolSignal is an explain-only tool for most captures. The safety gate classifies workflows as safe, explain-only, blocked, or unknown. For risky captures (e.g., vehicle systems, access credentials), it stays in explanation-only mode and does not provide transmit or replay steps.
What AI models can be used with SmolSignal?
SmolSignal supports Ollama, GPT/OpenAI, DeepSeek, Qwen/DashScope, llama.cpp, vLLM, or any OpenAI-compatible endpoint for optional AI-powered explanations.
Is SmolSignal safe to use?
Yes. The safety gate is deterministic and operates independently of any AI model. Hard-blocked categories such as car keys, access credentials, doors/gates, and bypass/cloning workflows cannot be overridden even in Authorized Lab Mode.
How do I install SmolSignal?
Requires Node.js 20 or newer and npm 10 or newer. Run 'npm install' then 'npm run dev' to start the development server, which typically opens at http://localhost:5173.