Typedguard is a Python library that implements fail-closed governance for LLM agent actions. It validates actions by checking typed fields rather than attempting to model or interpret free-form prose, providing stronger protection against prompt injection and policy violations. It is designed for developers building secure LLM agents and autonomous systems.
In the LLM eval & observability space, typedguard takes a focused approach. Safely governing LLM agent actions without relying on unreliable natural language interpretation of policies. It is built as an open-source project for developers. typedguard is open source under the Apache-2.0 license. It runs on the command line and API.
Behind typedguard is Nirmal Jingar, and it first shipped in 2026. Development happens publicly on GitHub with 6 commits in the last 90 days. Key capabilities include Type Checking, LLM Governance, and Prompt Injection Protection.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match