Articles

Token Strategy — Efficiency, Balance, Aggressive, and StaffingAnalysis — 분석2026-06-16Saving tokens and burning them are both strategies. Run the same loop in three modes — efficiency, balance, aggressive — and allocate mode and people to the business situation and evolution stage. This piece organizes token usage as a technical choice.token-strategycost-controlstaffingOCLSoperations
The Data Flywheel — Making the Loop Improve, Not Just PersistAnalysis — 분석2026-06-16Money runs the loop and tokens power the compute, but what makes the next turn better than the last is the data the operation accumulates. Turn the flywheel with three assets — traces, evals, learning signals — and keep data as an asset outside the model to preserve sovereignty.data-flywheeldata-sovereigntytraceprivate-evalOCLS
The Business Economics of the Loop — Earning and Surviving on TokensAnalysis — 분석2026-06-16Controlling cost alone does not sustain the loop. Read the revenue structure as token unit cost and margin, and choose how to survive among three paths — self-sustaining, investment, ecosystem. This piece lays out each path's realistic problems and best- and worst-case scenarios in the vocabulary of reopt architecture's evolution stages.token-economyunit-economicsbusiness-modelsustainabilityOCLS
Model Sovereignty and Human Agency — Two Axes on the Four LayersAnalysis — 분석2026-06-15Models get swapped and compute gets cheap. So what you control is not the model but the learning loop you build on top of it, and what aims that loop is human direction. This piece translates both axes into design using the vocabulary of reopt architecture's four layers, eight patterns, and OCLS loop.model-sovereigntyhuman-agencygovernanceOCLSprivate-eval
Running the reopt architecture five-phase loop in Claude CodeGuide — 가이드2026-04-17A practical guide that maps the Assess→Define→Loop→Execute→Evaluate methodology onto Claude Code's CLAUDE.md auto-load, slash commands, sub-agents, and skills — applied to a real repository.claude-codeworkflowOCLSgovernancemethodology
Implementing Reopt Agentic Governance: from assessment to systemCase Study — 사례 연구2026-04-15How a production SaaS running eight AI agents used reopt architecture's principles to diagnose its governance gaps and close them — universal audit logging, persistent approval records, MCP tracing and rate limiting, agent constraints, and an admin dashboard.reoptproductionauditapprovalMCPgovernanceguardrails
Building the Agent Catalog architecture with opt-harnessCase Study — 사례 연구2026-04-14A practical case study of using @reopt-ai/opt-harness, the design-governance harness package, to translate the Agent Catalog's architecture principles into a real product implementation.opt-harnessarchitecturegovernanceMCPrecipes