maestro

Use when bootstrapping, updating, or reviewing AGENTS.md — teaches what makes effective agent memory, how to structure sections, signal vs noise filtering, and when to prune stale entries

Use before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes

Deep discovery and specification for ambitious features. Full BMAD-inspired interview with classification, vision, journeys, domain analysis, and FR synthesis. Same output contract (spec.md + plan.md) as a standard feature but far richer. Use for multi-component systems, regulated domains, or unclear requirements.

Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies

Use when working with Docker containers — debugging container failures, writing Dockerfiles, docker-compose for integration tests, image optimization, or deploying containerized applications

Execute feature tasks following TDD workflow. Single-agent by default, --team for parallel Agent Teams, Sub Agent Parallels. Use when ready to implement a planned feature.

Create a new feature/bug track with spec and implementation plan. Interactive interview generates requirements spec, then phased TDD plan. Use when starting work on a new feature, bug fix, or chore.

Strategic analysis of a project to identify the single highest-leverage, most innovative addition. Use when the user asks what to build next, what the most impactful improvement would be, what's missing, or any question about strategic direction and priorities. Also use when stuck choosing between competing features.

Capture decisions, constraints, and context to persistent memory. Global memory is injected into every session and implementation run. Per-feature memory tracks working context.

Use when you need parallel, read-only exploration with task() (Scout fan-out)

Deep-review any plan (maestro, Codex, Claude Code plan mode, or plain markdown) using iterative subagent review loops with BMAD-inspired adversarial edge-case discovery. Spawns reviewer subagents that find issues using pre-mortem, inversion, and red-team techniques, auto-fixes them with structured fix strategies, and re-reviews until the plan passes with zero actionable issues. Use when the user says 'review the plan', 'deep review', 'check the plan thoroughly', 'review loop', 'validate before approving', or wants rigorous plan validation before execution. Also use proactively before plan-approve when the plan is complex or high-risk.

Strengthen a raw user prompt into an execution-ready instruction set for Claude Code, Amp, Codex, or another AI agent. Use when the user wants to improve an existing prompt, build a reusable prompting framework, wrap the current request with better structure, add clearer tool rules, or create a hook that upgrades prompts before execution.

Git-aware revert of feature, phase, or individual task. Safely undoes implementation with task state rollback.

Code review for a feature against its spec and plan. Verifies implementation matches requirements, checks code quality and security.

Review changed code for reuse, quality, and efficiency, then fix issues found. Use after implementing a task or feature -- catches duplication, hacky patterns, and wasted work before review.

Use when implementing any feature or bugfix, before writing implementation code

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always

Create interactive HTML visualizations of maestro state and debug data