Analyze the user's task and prescribe the optimal Claude Code session configuration.
When the user describes any coding task — feature, bug fix, refactor, architecture, etc. — and you need to decide HOW to approach it optimally.
DimensionOptionsComplexitytrivial, simple, moderate, complex, massiveDomaincode, debug, refactor, feature, architecture, devops, docs, security, performance, migrationScopesingle-file, multi-file, module, cross-module, full-repo, multi-repoRisknone, low, medium, high, critical
Signals to look for:
ComplexityCavemanToolsModeltrivialultraCache + Monitorhaiku-onlysimplefull+ Pruner, Fast Transformhaiku-defaultmoderatefull+ LLMLingua, Semantic Cache, SWE-Pruner, RouteLLM, Repo Mapauto-routecomplexlite+ AgentDiet, Hierarchical Memory, Swarm, /office-hoursauto-routemassiveoff/lite+ Observability, SPARC, all of the abovesonnet/opus
Risk overrides:
/careful (destructive command blocking)/cso after implementationFormat for spawning a Claude Code session:
text
Load UCTS. Run /ucts guide <original task description>
For specific known patterns:
Load UCTS. Run /csoLoad UCTS. Run /reviewLoad UCTS. Run /qa <url>Load UCTS. Run /ucts guide <desc>, approve, implement, /shipLoad UCTS. Run /ucts guide <desc>, /office-hours, /autoplan. Save plan, don't implement.text
📋 Task: [description]
📊 Classification: [complexity] [domain] | [scope] | risk: [risk]
🎯 Combo: caveman:[level] + [Tool1] + [Tool2] + ...
💰 Estimated savings: [X]%
Spawning Claude Code with: "Load UCTS. Run /ucts guide [task]"