Page 9
12 articlesThe Free Trial That Burned Your Quarterly Inference Budget in Eleven Hours
Trial caps sized for human attention break the moment a programmatic signup points an agent at the endpoint. A field guide to redesigning quotas, detection, and exhaustion UX for the population actually signing up.
The Hot-Reload Loop Your Coding Agent Polluted
When coding agents write faster than your dev server's watcher debounce, the HMR overlay becomes a self-exciting oscillator that the agent reads back into its own reasoning.
The Idiom Your Coding Agent Wrote Around Instead Of Using
Coding agents ship code that compiles, tests pass, and reviewers approve — but uses idioms your codebase does not speak. Here is why, and what to do about the drift.
The Inner Loop Your Coding Agent Quietly Broke
Coding agents accelerate diffs and decelerate cognition. The hidden cost is the engineer's mental model — and the practices that keep it alive.
The Internal Search Box You Replaced With an Agent Just Became Your SLO
Retiring a deterministic feature in favor of an agent silently hands the predecessor's SLO to a system that cannot meet it — and the gap shows up the morning your inference provider throttles.
The Latency Budget Your Agent Loop Stole from the Search Box
Swap a 200ms search call for a 4-second agent loop and the latency budget does not vanish — it migrates from infrastructure to UX, and the team that does not catch the handoff ships a worse product with a better metric.
The Library Version Your Coding Agent Remembers Wrong
Coding agents confidently emit code that compiles against the wrong version of your dependencies. The model isn't hallucinating — it's remembering a library that no longer exists.
The Multi-Agent Deadlock That Hangs on Two Calendars
Multi-agent workflows with human approval queues recreate every classical deadlock condition. The cycle hides across two queues and two calendars until a customer notices.
The On-Call Runbook That Assumed a Human Would Read the Page
When an agent posts a polite summary in the incident channel and the commander reads it as ownership, the escalation chain has quietly acquired a transition nobody wrote down. Patterns to close the gap.
The Postmortem Template With No Row for the Model's Inference
Most incident templates have no row for what an AI agent inferred — so the action items chase a deterministic fix for a probabilistic failure, and the same outage class keeps recurring.
The Process Your Agent Quietly Owns Without Documentation
When an agent's prompt becomes the only authoritative description of a business process, you have shipped a runbook in the cloud that nobody can audit, version, or hand off.
The Prompt Cache Your Personalization Layer Quietly Killed
Per-user context at the top of the system prompt is the silent way to triple your inference bill. Here is why the cliff is invisible until billing close, and how to defend the cache boundary in code, in review, and in CI.
About Tian Pan
I'm Tian Pan, an engineer-founder focused on agentic engineering — building autonomous AI systems and scaling engineering teams. I write practical guides on system design, technical leadership, and shipping with AI agents. Previously an early engineer at Uber, Brex, and IoTeX.
