agents

• How I Chained TX SOS and USPTO to Verify Entity Names Before FilingJuly 15th, 2026
• Why I Almost Rehomed My Entire Agent Fleet on Omnigent (And Why I Didn't)June 23rd, 2026
• The Retry Loop That Saved My 8 AM CEO Briefing (After Three Days of Silence)June 21st, 2026
• When Two AI Agents Write to the Same Memory at the Same TimeJune 17th, 2026
• Why I Stopped Saving Memory Every Turn (And What I Did Instead)June 17th, 2026
• A Fix the Agent Can't Run on Its Own Is Not a FixJune 11th, 2026
• The Silent-Fail Trap: Why HTTP 200 Doesn't Mean Your LLM Did AnythingJune 9th, 2026
• Every Document You Index Is Untrusted Input — Treat It That WayJune 4th, 2026
• When 'Make It All Searchable' Is the Wrong InstinctJune 3rd, 2026
• The settings.json Model Trap That Was Silently Burning My Opus BudgetMay 25th, 2026
• The Empty 200: What ZAI Taught Me About Silent Failures in Multi-Provider AI RoutingMay 24th, 2026
• How I Got Apollo to Call Me Back From AnywhereMay 14th, 2026
• Why I Budget Fifteen Minutes After Every Overnight Agent BuildMay 12th, 2026
• The Transient State That Ate Two Weeks of Fathom TranscriptsMay 11th, 2026
• The Dead-Code Fallback That Made Me Look Sideways for DataMay 10th, 2026
• I Built a Five-Phase AI Agent. It Died at the First Reboot.May 9th, 2026
• What Killing the Process Taught Me About Agent MemoryMay 6th, 2026
• The Stateless Agent Pattern: One Process Per Message, No ExceptionsMay 5th, 2026
• How I Learned to Block the Call, Not Just Filter the OutputApril 30th, 2026