Syntax - Tasty Web Development Treats 987: Remote Coding Agents
103 snips
Mar 16, 2026 Remote coding agents take center stage, from overnight grunt work and research tasks to travel planning and auto-fix workflows triggered by errors. There’s a tour of where these agents can live, including cloud tools, VPS setups, and dusty laptops at home. They also dig into browser control, SSH access, remote dev environments, and the chaos of API keys and permissions.
AI Snips
Chapters
Transcript
Episode notes
Wes Automated Tip Publishing Across Social Platforms
- Wes Bos built an agent skill to add tips to his website by gathering social links, captions, repo code, and supporting text.
- It automates a batching chore he used to avoid for months because manually collecting links across platforms was annoying.
Scott Turns Research Into Private AI Podcasts
- Scott Tolinski uses OpenClaw to research topics, summarize repos and blog posts, and turn them into private text-to-speech podcast episodes.
- He also uses similar agent workflows for travel planning, but stresses the results still need validation and verification.
Sentry Can Trigger Cursor To Fix New Errors
- Wes Bos wired Sentry to Cursor so new production errors automatically become prompts with metadata, codebase access, and possible pull requests.
- They frame it as self-healing software, while noting token tracking matters because an agent can spend real money on tiny fixes.