using-rune

Multi-agent engineering orchestration for Claude Code. Plan, work, review, inspect, test, and audit with Agent Teams. Includes 96 agent definitions (prompt templates — max 8 concurrent runtime agents via Agent Teams): 36 review, 5 research, 24 investigation, 19 utility, 6 work, 6 testing. 15 commands, 54 skills. Feature sediment detection with sediment-detector agent (SDMT-prefix) and validate-plugin-wiring.sh pre-commit script. UX Design Intelligence with 4 conditional UX review agents (heuristic, flow, interaction, cognitive) + ux-pattern-analyzer utility + ux-design-process skill (greenfield/brownfield methodology). Arc Phase 5.3 UX Verification integration. Curated elicitation methods for structured reasoning. Stack-aware intelligence with 12 language/framework/pattern specialist prompt templates (loaded on-demand). Figma design sync with VSM-based fidelity scoring. Inscription Sharding for parallel large-diff review. Multi-language quality guidance (Python, TypeScript, Rust, PHP, Axum). Talisman configuration mastery with stack-aware scaffolding. Session self-learning with CLI correction and review recurrence detection. Standalone browser E2E testing with human gate handling and interactive failure recovery. UntitledUI official MCP integration with builder-protocol metadata. Team Management SDK for centralized agent team orchestration.