Nunchuk has released two open source repositories that aim to reshape how AI agents interact with Bitcoin wallets, offering a model that limits agent control while maintaining human oversight.
the He releases Nunchuk includes the CLI, a command-line interface for managing Bitcoin wallets, and an accompanying “Agent Skills” repository designed to help AI systems run the CLI via common workflows. Both tools are licensed under MIT and target developers building automated financial systems on Bitcoin, Nunchuk said.
The basic premise challenges the growing trend in AI portfolio design. Instead of granting Agents have full control For funds with underlying collateral, Nunchuk proposes a shared custody model where agents operate within strict political boundaries. Human users retain final authority over transactions that exceed pre-defined limits.
Under this structure, wallets are configured as collective wallets with multiple keys. The user key, proxy key, and policy co-signer work together to allow transactions. The agent can initiate actions such as creating wallets, inviting participants, and creating transactions, but spending authority remains limited by rules established at the policy level.
Using Bitcoin based on artificial intelligence
These policies set limits such as daily spending caps, approval requirements, and signature delays. Transactions that fall within the permitted parameters can proceed without intervention, while larger or sensitive actions require explicit user consent.
Nunchuk separates incubation from automation. The wallet structure governs ownership and control of funds, while policy layers determine what the agent can execute. This distinction ensures that financing a portfolio does not give broader power to the agent managing it.
The CLI supports a range of functions including key generation, wallet creation, transaction workflow, and policy configuration. It also allows users to export wallet descriptors and backups using standard formats, which supports portability and recovery outside the Nunchuk ecosystem.
The agent skill repository serves as an interface layer for AI systems. It provides predefined commands and prompts to guide agents through tasks such as setting up portfolios, managing participants, and executing transactions. This reduces the need for custom integrations and lowers the barrier Developers are experimenting with Bitcoin-based automation.
Nunchuk positions the dual warehouse approach as a response to two distinct challenges: implementation and ease of use. The command line interface acts as an implementation layer linked to the Nunchuk API, while the skills layer focuses on how AI systems interact with that infrastructure.
The release reflects a broader effort to identify safe design patterns for AI in financial contexts. By enforcing constrained authority, Nunchuk aims to enable practical automation without introducing the risks of full custody.
Potential use cases include shared wallets for human agents, automated bill payment systems, treasury management tools, and coordination between multiple agents. While these applications are still in their early stages, the basic model provides a framework for controlled experiments.
As AI systems gain access to financial instruments, the issue of control becomes central. The Nunchuk approach suggests that the path forward may depend less on constraining capabilities and more on structuring power.
Editorial disclaimer: We leverage AI as part of our editorial workflow, including supporting research, image generation, and quality assurance processes. All content is directed, reviewed and approved by our editorial team, responsible for accuracy and integrity. AI-generated images only use tools properly trained on licensing materials. In Bitcoin, as in the media: don’t trust. Confirms.
