Cloudflare Security Audit Skill Review: Agent Bug Hunting
Cloudflare's security-audit skill runs 6-phase audits with isolated hunters and adversarial verifiers. 17K stars, MIT. Setup, JSON schema, token cost, limits.
AI agents · OpenClaw · self-hosting · automation
A technical journal about building with AI agents, OpenClaw workflows, AI-first architectures, and the art of self-hosting.
Written by humans. Optimized for AI discovery.
Cloudflare's security-audit skill runs 6-phase audits with isolated hunters and adversarial verifiers. 17K stars, MIT. Setup, JSON schema, token cost, limits.
shadcn/lint enforces Tailwind design-system rules with errors that tell agents what to use instead. 2.1K stars in 5 days, ESLint + Oxlint, MIT. Evals + limits.
OpenBot gives each AI agent its own browser, files and shell behind a policy gateway that audits every action. MIT, self-hosted, any AG-UI agent. Setup, limits.
WeKnora is Tencent's MIT-licensed knowledge platform: RAG Q&A, a ReAct agent with sandboxes, and a self-maintaining wiki. Docker install, API code, limits.
Google's ARTEMIS drives real Android phones from Claude Code, Codex and Antigravity via MCP. 6k stars, 99% AndroidWorld — and the Minitap attribution dispute.
Direct answers to the most-asked AI questions. Updated daily.