# Universal Autonomous Agent Policies

A library of copy-paste prompts, each defining a specialized autonomous agent for one area of codebase maintenance, security, performance, testing, documentation, databases, and more, meant to be pasted into any AI agent platform for an arbitrary codebase.

Universal Autonomous Agent Policies is a set of portable, specialized prompts rather than installable software: each prompt separates a specialist policy, a repository adapter, and an execution step, and adapts only to capabilities it can actually verify in the target codebase rather than assuming they exist. Every agent inspects the git and repository state first, discovers the actual languages, frameworks, build and test tooling, and conventions in use, classifies each finding as detected, not detected, or unknown, and reports not applicable when its specialty does not apply.

The collection includes over twenty specialist prompts covering areas such as application security, performance, defect hunting, test quality, accessibility, documentation, databases, APIs, observability, CI/CD, Docker, Kubernetes, Terraform, mobile, and machine-learning systems, each following the same orient, discover, adapt, baseline, prioritize, implement, verify, review, and document lifecycle. A governing document defines this contract and the boundaries between agents, and a prompt is used by copying it directly into an AI agent platform.

Website: https://julesai.pages.dev/