<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>dbward blog</title><description>Technical articles about database operations, approval workflows, and AI agent safety.</description><link>https://dbward.dev/</link><item><title>Slack approvals for production SQL — without exposing SQL</title><link>https://dbward.dev/blog/slack-approval/</link><guid isPermaLink="true">https://dbward.dev/blog/slack-approval/</guid><description>Most approval workflows fail at the context switch. dbward lets reviewers inspect SQL, EXPLAIN, and risk in a Slack modal while keeping queries out of channels.</description><pubDate>Thu, 09 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Deploying a centralized DB approval system without opening inbound ports to your databases</title><link>https://dbward.dev/blog/agent-architecture/</link><guid isPermaLink="true">https://dbward.dev/blog/agent-architecture/</guid><description>Most DB management tools require the control plane to connect to databases. That architecture turns network boundary management into the primary adoption cost. dbward&apos;s agent model inverts this: credentials and database connectivity stay local, only outbound HTTP(S) leaves.</description><pubDate>Sun, 28 Jun 2026 00:00:00 GMT</pubDate></item><item><title>AI agents should not have direct production DB credentials</title><link>https://dbward.dev/blog/ai-agent-credentials/</link><guid isPermaLink="true">https://dbward.dev/blog/ai-agent-credentials/</guid><description>MCP made it easy to connect AI agents to databases. But &apos;easy to connect&apos; isn&apos;t &apos;safe to connect.&apos; Here&apos;s how to build the execution boundary that&apos;s missing.</description><pubDate>Thu, 25 Jun 2026 00:00:00 GMT</pubDate></item></channel></rss>