Show HN: Aigenos Flow – deterministic agents on LangGraph; the policy is a file
Category: ai-ml
Tags: langgraph, agent-workflows, policy-enforcement, deterministic-execution, llm-security, human-in-the-loop
Score: 7.8/10 (Innovation: 8, Technical: 8, Documentation: 8, Utility: 7)
Aigenos Flow is a deterministic agent framework that enforces workflow policies declared in a single SKILL.md file, using LangGraph as the execution backend. It prevents LLMs from bypassing critical steps like confirmations or 2FA by making the policy file the source of truth and rejecting invalid skills before runtime. The project offers a clean separation between LLM-driven natural language understanding and rule-based execution, with typed widgets, audit trails, and human-in-the-loop approvals.
Target audience: backend developers building LLM agents, AI engineers, security-conscious devops
Repository: https://github.com/aigenos/flow · Python · Apache-2.0 · 1 stars
View on Hacker News