Structured divergence engine. Generates a wide, deliberately non-anchored set of approaches to a consequential, open-ended problem by spawning isolated parallel sub-agents under different cognitive frames, then a separate critic pass scores, declusters, and deepens the best. The divergent complement to /deep-analysis. Use for high-stakes architecture, strategy, design-space, and "what are our options" decisions. Auto-activates on diverge, brainstorm options, explore alternatives, what are our options, de-anchor, divergent thinking, option space, consider alternatives, multiple approaches.
Orchestrated task execution engine. Decomposes any goal into small atomic tasks, plans dependencies, selects the right agent/tool/MCP server for each, executes in optimally parallel batches, and tracks everything. Use when given a complex, multi-step goal that benefits from structured decomposition and full tool utilization.
Advanced multi-agent patterns using Claude Code's built-in orchestration. Teams, background agents, coordinator pattern, worker restrictions, SendMessage protocol, fork subagents. Auto-activates on multi-agent, orchestrate, team, coordinator, parallel agents, worker agents, background agent, agent swarm.
Structured multi-step reasoning for complex problems using the sequential-thinking MCP server. Use when facing architectural decisions, performance bottlenecks, complex debugging, design trade-offs, technology selection, or any problem requiring rigorous step-by-step analysis with hypothesis testing.
Deep root cause analysis engine. Systematically investigates bugs, crashes, unexpected behavior, and performance issues through an 8-phase diagnostic protocol. Uses structured reasoning (sequential-thinking MCP), multi-pass codebase analysis, git forensics, evidence-based hypothesis testing, and the 5 Whys method. Never jumps to a fix — always proves the root cause first.
Comprehensive codebase reading engine. Systematically reads actual source code line by line through a 6-phase protocol — scoping, structural mapping, execution tracing, deep reading, pattern synthesis, and structured reporting. Source code is the source of truth. Use when needing to truly understand how code works, not just what documentation claims.
Explicit-level policy for context management in long-horizon Claude Code sessions. Five escalating responses to context pressure — apply the cheapest intervention that still preserves the work. Auto-activates on context limit, near context limit, conversation getting long, /compact, handoff, /save-session.
E2E browser testing with Playwright MCP. Visual regression, responsive design, cross-browser testing, codegen sessions. Auto-activates on e2e test, visual regression, screenshot test, browser test, Playwright, responsive test, cross-browser, visual verification.