https://h5i.dev/blog/ Design essays on giving an AI agent a browser you can audit: boundaries, evidence, and what a request log has to prove. en-usSun, 13 Sep 2026 12:00:00 GMT https://h5i.dev/blog/ai-pentesting-tools/ https://h5i.dev/blog/ai-pentesting-tools/Wed, 09 Sep 2026 12:00:00 GMT Burp Suite, ZAP, Caido, and h5i compared for AI pentesting: AI-assisted manual testing, AI-authored scan automation, and agent-led investigation. https://h5i.dev/blog/the-h5i-loop/ https://h5i.dev/blog/the-h5i-loop/Fri, 21 Aug 2026 12:00:00 GMT Create one sandbox for an AI coding task, browse from inside it, develop and test there, then review the evidence before exporting or applying the patch. https://h5i.dev/blog/the-environment-is-the-sandbox/ https://h5i.dev/blog/the-environment-is-the-sandbox/Fri, 21 Aug 2026 12:00:00 GMT AI coding agents run package scripts, compilers, tests, servers, and browsers. Sandboxing the whole workload limits what mistakes and untrusted code can reach. https://h5i.dev/blog/choosing-agent-isolation/ https://h5i.dev/blog/choosing-agent-isolation/Fri, 21 Aug 2026 12:00:00 GMT Choose an AI coding-agent sandbox by the failure it must prevent: host file access, unrestricted network traffic, environment drift, or a shared kernel. https://h5i.dev/blog/evidence-for-agent-work/ https://h5i.dev/blog/evidence-for-agent-work/Fri, 21 Aug 2026 12:00:00 GMT Review AI-generated code using the diff, externally observed test results, denied actions, browser errors, and explicit gaps in evidence. https://h5i.dev/blog/prompt-injection-is-a-boundary-problem/ https://h5i.dev/blog/prompt-injection-is-a-boundary-problem/Fri, 21 Aug 2026 12:00:00 GMT Limit the files, credentials, network destinations, local services, and repository writes available to a prompt-injected coding agent. https://h5i.dev/blog/burp-suite-vs-h5i-for-ai-agents/ https://h5i.dev/blog/burp-suite-vs-h5i-for-ai-agents/Wed, 09 Sep 2026 12:00:00 GMT Burp Suite vs h5i for AI agents: compare HTTP capture, request editing, browser automation, and fully automated penetration testing with Burp AT or h5i. https://h5i.dev/blog/owasp-zap-vs-h5i-for-ai-agents/ https://h5i.dev/blog/owasp-zap-vs-h5i-for-ai-agents/Wed, 09 Sep 2026 12:00:00 GMT OWASP ZAP vs h5i for AI agents: compare DAST scanning and MCP automation with fully automated, agent-driven browser testing, HTTP capture, editing, and replay. https://h5i.dev/blog/caido-vs-h5i-for-ai-agents/ https://h5i.dev/blog/caido-vs-h5i-for-ai-agents/Wed, 09 Sep 2026 12:00:00 GMT Caido vs h5i for AI agents: compare proxy history, HTTPQL, Replay, Automate, and Skills with fully automated agent-driven browser and HTTP testing.