Products
Kill Switch: Remote Kill Switch for AI and AI Agents
launched
A safety middleware that screens LLM inputs and outputs and lets operators remotely disable model or agent usage at any scope. Rules are written in plain English and enforced via two API calls wrapping the LLM. It targets developers wanting guardrails against runaway or dangerous AI agent behavior.
Visit project ↗4 engagement1 sourceSat, Oct 3, 2026, 05:44 PM
AI Overview
- Screens LLM inputs and outputs, enabling remote shutoff of AI/agent use at broad or narrow scope
- Rules defined in plain English
- Integrates with two calls around the LLM; adds ~90–300ms latency
- Built by a post-quantum cryptography practitioner and early AI agent OSS contributor
Related
Products·building
AgentGuard — permission layer that gates AI agent tool execution
AgentGuard is a small permission layer that sits between an AI agent and its tools, checking each call before execution. It verifies tool allowance in the current agent state, argument policy compliance, and unknown/unsafe states, and ensures denied functions remain untouched. It targets developers building LLM agents who need guardrails on tool use.
AgentGuard is a small permission layer that sits between an AI agent and its tools, checking each call before execution. It verifies tool allowance in the current agent state, argument policy compliance, and unknown/unsafe states, and ensures denied functions remain untouched. It targets developers building LLM agents who need guardrails on tool use.
7 engagement1 sourceredditWed, Sep 2, 2026, 10:26 AM
Products·launchedfeatured
Safe execution framework for LLM tool calls with mechanical guarantees
A framework that enforces mechanical safety guarantees on LLM tool calls, preventing irreversible actions like unauthorized payments or database deletions. It solves the problem of blind trust in AI agents for developers and enterprises deploying LLMs with tool access.
A framework that enforces mechanical safety guarantees on LLM tool calls, preventing irreversible actions like unauthorized payments or database deletions. It solves the problem of blind trust in AI agents for developers and enterprises deploying LLMs with tool access.
121 engagement1 sourceyoutubeMon, Jul 13, 2026, 07:25 PM
Products·launched
Attestor: open-source zero-trust boundary for autonomous AI agents
Attestor is an open-source execution boundary that enforces zero-trust security for autonomous AI agents. It uses LLMs to define and enforce policies that restrict agent actions, preventing unauthorized access or data leaks. This solves the problem of safely deploying autonomous agents in production environments.
Attestor is an open-source execution boundary that enforces zero-trust security for autonomous AI agents. It uses LLMs to define and enforce policies that restrict agent actions, preventing unauthorized access or data leaks. This solves the problem of safely deploying autonomous agents in production environments.
3 engagement1 sourcehackernewsSun, Jul 12, 2026, 07:54 AM
Products·launched
PromptSentry: 3-layer LLM security proxy with sub-ms prompt injection blocking and local DLP scrubbing
PromptSentry is a FastAPI reverse proxy that sits between an app and its LLM endpoint, filtering requests through three layers. It uses regex/AST heuristics to block prompt injections in under a millisecond and local DLP to redact sensitive data before it leaves the network. It targets developers who need to secure LLM API traffic without adding latency or sending secrets to third parties.
PromptSentry is a FastAPI reverse proxy that sits between an app and its LLM endpoint, filtering requests through three layers. It uses regex/AST heuristics to block prompt injections in under a millisecond and local DLP to redact sensitive data before it leaves the network. It targets developers who need to secure LLM API traffic without adding latency or sending secrets to third parties.
0 engagement1 sourceredditMon, Sep 28, 2026, 05:58 AM
Products·launched
Doberman: AI watchdog that blocks dangerous LLM commands
Doberman is a two-layer security system that monitors LLM inputs, outputs, and tool calls to prevent dangerous actions like database deletion. It combines deterministic rules with a dynamic layer that learns user preferences, addressing the gap in existing guardrails that miss 11% of threats.
Doberman is a two-layer security system that monitors LLM inputs, outputs, and tool calls to prevent dangerous actions like database deletion. It combines deterministic rules with a dynamic layer that learns user preferences, addressing the gap in existing guardrails that miss 11% of threats.
11 engagement2 sourceshackernews, redditMon, Aug 17, 2026, 08:03 PM