Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.joggr.ai/llms.txt

Use this file to discover all available pages before exploring further.

Before Guided Generative Workflows (GG) can research or plan effectively, it needs to understand your codebase — your architecture, conventions, stack, integrations, and testing patterns. Rather than re-analyzing the repository from scratch on every project, Workflows generates a set of codebase documentation files once and reuses them across all projects. These files feed into /gg-research, /gg-plan, and /gg-execute for every project you run. This also keeps token usage low — instead of crawling your entire repo on every project, Workflows reads these pre-generated files and focuses only on what is unique to the current task. Run /gg-codebase once when you set up Workflows in a repository. Every subsequent project can reference these files — you do not need to re-research your codebase for each new project.

​
What Gets Generated

/gg-codebase writes six files to .joggr/.gg/codebase/:
FileWhat it captures
architecture.mdCodebase structure, layers, component interactions, and where to add new code.
conventions.mdCode style, naming conventions, directory structure, and best practices.
development.mdCommit conventions, branch naming, CI/CD pipeline, and development rules.
integrations.mdExternal services, APIs, and third-party dependencies.
stack.mdLanguages, frameworks, databases, and tools.
testing.mdTest framework, test structure, patterns, and coverage.
These files are committed to your repository and shared across your team. When /gg-research runs for a new project, it reads these files instead of crawling the repo again — keeping research focused on what is unique to the current task.

​
Focus Areas

You can limit analysis to a specific area: 
/gg-codebase tech      # stack.md + integrations.md
/gg-codebase arch      # architecture.md
/gg-codebase quality   # conventions.md + testing.md
/gg-codebase dev       # development.md
Use a focus area to quickly refresh a specific section — for example, after adding a new dependency, run /gg-codebase tech to update just the stack and integrations docs.

​
When to Re-run

The codebase files are stable until your repository structure changes significantly. Re-run /gg-codebase after:
  • A major refactor or restructured directory layout
  • Adding a new framework or replacing a core dependency
  • Changing your CI/CD pipeline or development workflow
After analysis, Workflows presents a summary of what was written and asks you to review the output. If anything looks wrong, describe the issues and it will re-analyze the affected areas.