Results for “inference-latency”

Orchestrate token-heavy research, coding, and testing by delegating bounded tasks to cheaper subagents while reserving Claude Fable for architecture, synthesis, and final review.

Diagnoses why someone knows what to do but doesn't do it, using Adlerian psychology to identify root causes like procrastination, perfectionism, or avoidance.

Diagnoses whether a user's current approach is trading short-term speed for long-term pain, and recommends slower methods that build compoundable assets through friction.

Debug and optimize AI agents by analyzing reasoning traces, context degradation, tool confusion, instruction drift, repeated task failures, and performance regressions.