Show HN: Heides, a deterministic code harness giving AI agents senses and judgment
Sonde builds a code graph locally that AI agents can query for accurate structural facts about a codebase instead of relying on LLM guesses. It targets developers building coding agents who need grounded, verifiable context. The LLM consumes the graph output rather than generating it.
Sonde builds a code graph locally that AI agents can query for accurate structural facts about a codebase instead of relying on LLM guesses. It targets developers building coding agents who need grounded, verifiable context. The LLM consumes the graph output rather than generating it.
A fail-closed evidence engine that helps AI agents comply with norms by providing verifiable evidence. It uses LLMs to interpret norms and generate evidence, ensuring agents act within defined boundaries. Solves the problem of AI agents acting unpredictably in regulated environments.
A fail-closed evidence engine that helps AI agents comply with norms by providing verifiable evidence. It uses LLMs to interpret norms and generate evidence, ensuring agents act within defined boundaries. Solves the problem of AI agents acting unpredictably in regulated environments.
Heimdall is a knowledge layer that provides AI coding agents with verified, trustworthy information to improve their outputs. It uses LLMs to process and verify knowledge, solving the problem of unreliable information in AI-assisted coding. Aimed at developers using AI coding agents.
Heimdall is a knowledge layer that provides AI coding agents with verified, trustworthy information to improve their outputs. It uses LLMs to process and verify knowledge, solving the problem of unreliable information in AI-assisted coding. Aimed at developers using AI coding agents.
Stonefold is a gateway that provides deterministic control for AI agents interacting with external systems. It uses LLMs to interpret user requests but enforces predefined workflows and validations to ensure reliable outcomes. It solves the problem of unpredictable agent behavior for enterprises needing safe automation.
Stonefold is a gateway that provides deterministic control for AI agents interacting with external systems. It uses LLMs to interpret user requests but enforces predefined workflows and validations to ensure reliable outcomes. It solves the problem of unpredictable agent behavior for enterprises needing safe automation.
A developer introduced AletheionAGI, a tool for enforcing grounding in AI agents, on Hacker News. The tool aims to keep agent outputs tied to verified sources.
A developer introduced AletheionAGI, a tool for enforcing grounding in AI agents, on Hacker News. The tool aims to keep agent outputs tied to verified sources.