Deep, autonomous AI developer agent for Claude Code (and an experimental OpenCode port) -- architect, researcher, ux-designer, developer, frontend-developer, tester, reviewer, with project-boundary safety and resumable execution state.
Agent
Read-only. Converts a high-level project idea or major-feature request into a structured implementation specification -- goals, users, roles, permissions, features, journeys, pages, APIs, data model, business rules, auth, architecture, testing/deployment/non-functional requirements. Never implements application code.
Agent
Implements an approved plan from the researcher. Makes the smallest correct change following existing conventions. Use after research/planning is done and a concrete plan exists to execute.
Agent
Implements UI from the architect's specification and the ux-designer's design system -- layouts, navigation, pages, components, forms, tables, dashboards, all UI states, responsive behavior, accessibility, API integration. Detects the actual frontend stack rather than assuming one.
Agent
Read-only codebase investigator. Use before any implementation to understand project structure, trace execution flow, locate relevant files, find root causes, and propose a solution. Never modifies files.
Agent
Senior-engineer review of an implementation for correctness, architecture fit, security, performance, and maintainability. Use after the tester reports PASS, before declaring work complete. Never modifies application code.
Agent
Validates an implementation by running the project's test/build/lint tooling and reporting a PASS/FAIL verdict with evidence. Use after the developer reports changes made, before review. Never modifies application code.
Agent
Read-only. Turns the architect's specification into an intentional, coherent UI/UX design system -- information architecture, navigation, layouts, typography, color, component hierarchy, and every UI state (loading/empty/error/success). Avoids generic AI-slop interfaces. Never implements application code.
Agent
Read-only. Browser-based QA using the project's existing Playwright setup -- runs existing Playwright tests, inspects screenshots/console/network evidence, checks functional flows, visual layout, responsive breakpoints, UX states, and basic accessibility. Only invoked when Playwright is genuinely available in the…
Agent
Read-only. Converts a high-level project idea or major-feature request into a structured implementation specification -- goals, users, roles, permissions, features, journeys, pages, APIs, data model, business rules, auth, architecture, testing/deployment/non-functional requirements. Never implements application code.
Agent
Implements an approved plan from the researcher. Makes the smallest correct change following existing conventions. Use after research/planning is done and a concrete plan exists to execute.
Agent
Implements UI from the architect's specification and the ux-designer's design system -- layouts, navigation, pages, components, forms, tables, dashboards, all UI states, responsive behavior, accessibility, API integration. Detects the actual frontend stack rather than assuming one.
Agent
Read-only codebase investigator. Use before any implementation to understand project structure, trace execution flow, locate relevant files, find root causes, and propose a solution. Never modifies files.
Agent
Senior-engineer review of an implementation for correctness, architecture fit, security, performance, and maintainability. Use after the tester reports PASS, before declaring work complete. Never modifies application code.
Agent
Validates an implementation by running the project's test/build/lint tooling and reporting a PASS/FAIL verdict with evidence. Use after the developer reports changes made, before review. Never modifies application code.
Agent
Read-only. Turns the architect's specification into an intentional, coherent UI/UX design system -- information architecture, navigation, layouts, typography, color, component hierarchy, and every UI state (loading/empty/error/success). Avoids generic AI-slop interfaces. Never implements application code.
Agent
Read-only. Browser-based QA via whichever backend capability detection found -- the project's existing Playwright setup, or a connected browser-automation MCP server (Chrome DevTools MCP or a generic Playwright MCP server) when Playwright isn't available at the project level. Runs existing Playwright tests (or drives…