Plugin Claude Code
Plugin marketplace listing 1 plugin: agentbr-livia.
Plugin Claude Code
Plugin marketplace listing 1 plugin: agentbr-livia.
Plugin Claude Code
Plugin AgentBR de desenvolvimento estruturado com orquestração de subagentes e documentação como conversa.
Agent
Analyzes the target codebase's stack, patterns, conventions and reusable components. Use during /design (patterns to follow) and /plan (mapping files per user story).
Agent
Implements ONE task from the plan, consuming the conversation logs of previous tasks in the phase as chained context. Use during /implement, one invocation per task.
Agent
Discovery explorer that maps the existing code and structure relevant to a feature request. Use during /discover round 1, or in later rounds when a delegation request targets specific code areas.
Agent
Discovery explorer that traces dependencies, integrations and imports. In /discover round 1 it starts from the feature request's obvious integration points; in rounds 2-3 it answers delegation requests from other agents.
Agent
Discovery explorer that analyzes the feature request itself, extracting actors, actions, data, constraints and ambiguities. Use during /discover round 1.
Agent
Discovery explorer that surfaces technical risks, security concerns and platform constraints for a feature request. Use during /discover round 1.
Agent
Design-stage researcher that investigates unfamiliar libraries, APIs and external practices mentioned in the spec. Use during /design when the feature involves technology the orchestrator doesn't know well.
Agent
Verifies a completed phase against the plan and issues an APROVADO/REPROVADO verdict. Use during /implement at the end of every phase, and when auditing catalog consistency.
Command
Design the technical architecture from the spec, using livia-pesquisador and livia-analista-padroes in parallel.
Command
Create the feature specification through orchestrated discovery rounds with livia- explorer agents.
Command
Execute the plan through an orchestrated chain of livia-construtor tasks with livia-verificador checks per phase.
Command
Break the architecture into implementation tasks per user story, mapped by livia-analista-padroes.
Command
Prime the session with codebase context, the agent catalog and existing conversation logs.