9 specialized agents. 10 skill workflows. One command.
From zero to a structured, multi-agent development workflow in minutes.
Interactive onboarding. Generates 9 agents and 10 skill workflows tailored to your project.
Explores your codebase. Enriches CLAUDE.md with real context, conventions, and patterns.
Full pipeline: evaluation, spec, implementation, code review, and QA. End to end.
Each agent handles one responsibility. The pipeline moves forward automatically.
Each agent has a single responsibility and follows the project conventions defined in your CLAUDE.md.
Evaluates ideas and gives strategic direction. First gate before any work begins.
Converts ideas into concrete tasks with acceptance criteria and clear scope.
Defines technical approach, architecture, and implementation direction.
Implements features following project conventions, writes tests, makes atomic commits.
Reviews code quality, patterns, and tech debt. Guards the standards of the guild.
Testing, edge cases, and regression. Validates that the implementation meets acceptance criteria.
Diagnosis and bug resolution. Isolates root causes and applies targeted fixes.
Schema changes and safe migrations. Handles rollback strategies and data integrity.
Resolves Claude Code integration issues, worktree configs, and environment problems.
Claude Code without structure leads to inconsistent results. Guild gives every session a clear process: evaluate, plan, implement, review, test.
Agents define identity and expertise. Skills define workflows and steps. Clean separation that scales without complexity.
CLAUDE.md holds project conventions. SESSION.md tracks what happened. PROJECT.md keeps the roadmap. Context survives across sessions.
No servers, no databases, no API keys. Just markdown files and Claude Code. Works in any project, any language, any stack.