TACO: A Toolsuite for the Verification of Threshold Automata

TACO (Toolsuite for Verification of Threshold Automata) is a new framework for automatically verifying fault-tolerant and threshold-based distributed algorithms. The tool, developed by Paul Eichler, Tom Baumeister, Mouhammad Sakr, Mahboubeh Kalateh Dowlati, Marcus Völp, and Swen Jacobs, provides three model checking approaches that cover different decidable fragments of threshold automata, plus two semi-decision procedures that go beyond those fragments. This makes TACO one of the most comprehensive verification tools for threshold automata, capable of handling a wide range of distributed consensus and fault tolerance algorithms.

TACO is designed as a modular, extensible, and well-documented framework, allowing researchers and engineers to develop new verification algorithms or integrate existing ones. The paper evaluates TACO's performance experimentally, showing its practicality for real-world distributed systems. Published as an extended version of a CAV 2026 paper, TACO addresses a critical need: ensuring correctness of distributed algorithms that rely on thresholds (e.g., number of faulty nodes) to maintain system reliability. For professionals building fault-tolerant systems in cloud computing, blockchain, or critical infrastructure, TACO offers a rigorous way to prove algorithmic correctness automatically.