Jentic has launched Jentic Mini, a free, open-source, self-hosted tool intended to help developers control how general-purpose AI agents connect to APIs and workflows.
The Dublin-based company said the product is designed for developers building with OpenClaw, and aims to reduce risks associated with giving agents access to live systems, including credential exposure, excessive permissions and unintended actions in production environments.
Jentic Mini provides what the company describes as a “safety and control layer” around agent access, including fine-grained permissions and a single “killswitch” to shut down an agent’s data access. The release also connects developers to an AI-curated catalogue of more than 10,000 APIs and workflows, which Jentic said is updated over time.
The launch comes as organisations experiment with increasingly capable agentic systems that can automate tasks across internal and third-party services. Security teams have raised concerns that these agents often require broad access to tools and credentials, creating new pathways for misuse or compromise if access controls and governance are weak.
Jentic said its catalogue is intended to provide a structured “map” of tools and workflows that agents can use, supporting more controlled integration with real systems. The company also said Jentic Mini can be used with other general-purpose agents, including NemoClaw.
Jentic also noted it is available as a verified connector in Anthropic’s Claude, which it said allows users to access connected tools and APIs through the assistant.
Jentic Mini is available via the company’s website and on GitHub.
