claude-world

Un agente de investigación de IA totalmente autónomo que ingiere fuentes en Google NotebookLM, realiza investigación en la web profunda y sintetiza conocimiento a través de preguntas y respuestas citadas y 9 tipos de artefactos descargables.

A helpful skill that does many things

A simple skill that helps with greeting messages

TDD-based autonomous development loop with checkpoint recovery and observability changelog

Generate custom agent from template

Validate custom agent file format and structure

List all available agents (core + expert)

Track development session events in a daily markdown changelog, including file changes, test results, and key decisions.

View and manage the runtime changelog for observability

Generate CLAUDE.md template for current project

Documentation specialist for README, API docs, and code comments

Self-Evolving Development Loop - Dynamic skill generation with learning and evolution

Delegate tasks to Gemini CLI to save Claude context

Automatically routes tasks to external AI CLIs (Codex or Gemini) when more efficient. No manual commands needed - routing decisions are made automatically based on task type.

Verify development environment is ready

Debugging specialist for root cause analysis and problem resolution

View Self-Evolving Loop session status, history, and memory metrics

Guided 5-minute onboarding for Director Mode Lite

Generate hook script from template

Validate CLAUDE.md structure and completeness

Code review specialist for quality, security, and best practices

Problem analysis using Explore agents

Delegate tasks to Codex CLI to save Claude context

Validate hooks configuration and scripts

Validate skill/command file format and structure

Test-Driven Development (TDD Red-Green-Refactor)

Test automation specialist for running tests and ensuring coverage

Validate MCP configuration and suggest improvements

Expert-guided project setup with 6 phases

List all available skills (core + custom)

Complete 5-step development workflow

Create detailed execution plan with task breakdown

Complete project health audit (7 checks)

Generate custom skill/command from template

Conventional Commits with quality checks

Compress agents/skills with quality assurance loop. Iterative 5-phase process ensures understanding preserved. TRIGGERS: semantic-compressor, compress skill, optimize agent, reduce tokens, benchmark