#automation

4 posts

Claude Headless Sandboxing: Why /tmp Writes Fail in Non-Interactive Mode

Claude Headless Sandboxing: Why /tmp Writes Fail in Non-Interactive Mode

Claude Code in print mode denies writes outside its working directory with no prompt and no error. Here's why /tmp breaks headless pipelines and the fix.

Blog Pipeline Fence Corruption: How a Naive Sed Rule Swallowed Markdown Code Blocks

Blog Pipeline Fence Corruption: How a Naive Sed Rule Swallowed Markdown Code Blocks

A one-line sed cleanup rule stripped every closing code fence in LLM-generated posts. Here's the failure mode, the fix, and a fence validator that catches it.

When Agents Should Stop: Designing Safety Boundaries That Work

When Agents Should Stop: Designing Safety Boundaries That Work

Stop conditions are a feature, not a failure state. Budget caps, progress checks, and exit contracts that make autonomous agents quit cleanly.

Self-Improving AI Infrastructure: How Your Homelab Wiki Updates Itself

Self-Improving AI Infrastructure: How Your Homelab Wiki Updates Itself

How to automate your homelab wiki with self-improving AI infrastructure

← All tags