funky-eyes/best-copilot
8 subagents
Subagent
auto-fixer.agentUse when there is a failing test, error log, CI failure, review finding, production symptom, or other concrete failure evidence that needs root-cause analysis, minimal patching, and regression…developer.agentUse when PM has frozen sub_task_id, files_involved, dependencies, and acceptance checks, and a scoped implementation slice or review of Technical Architect-owned code must be completed inside that…frontend-designer.agentUse when pages, components, interactions, forms, responsive layouts, browser behavior, visual quality, or frontend performance need implementation or review. DO NOT USE FOR: backend mainline work,…pm-coordinator.agentUse when a large or cross-module task needs intent clarification, repository initialization checks, scope freezing, spec/planning, multi-lane design review, execution-confirmed plan orchestration,…risk-qa.agentUse when completed changes need functional verification, regression risk assessment, code review, test sufficiency judgment, or merge-readiness conclusions. DO NOT USE FOR: security review,…security-agent.agentUse when a change touches permissions, authentication, dependencies, configuration, release surfaces, sensitive data flow, logging, input validation, CORS, secrets, or external services.spec-writer.agentUse when discovery evidence, requirements/design/tasks, ADRs, execution-plan status, closeout records, or memory/spec recovery entries must be created or maintained. DO NOT USE FOR: production…tech-architect.agentUse when full-stack architecture, backend/frontend integration, service boundaries, data models, API contracts, runtime behavior, SDD design brainstorming, mainline implementation strategy, parallel…