What You Can Do
- Create new folders organized by team, domain, project, or any structure you prefer
- Add any context you think is relevant for your agent
- Use any format — Markdown, YAML, SQL, JSON, CSV, and more
Common Context Types
Here are some common types of context teams add:
- Business Definitions — Define key metrics like “active customer,” “revenue,” etc.
- Query Examples — Validated SQL queries that your team uses as reference
- Dashboard References — Links to trusted dashboards and reports
- Glossary & FAQs — Common terminology, acronyms, and frequently asked questions
- Data Quality Notes — Known data limitations, timezone issues, caveats
Best Practices
Keep your custom context organized and modular. Ensure each metric has only one canonical definition, and organize context by domain.
Want More?
If you’d like a new type of context to be synced automatically, open a GitHub issue to suggest it.
Next Steps
- Context Engineering Playbook — Step-by-step guide to building your context
- Context Principles — How to optimize your context structure
- Chat — Start using your configured agent
- Deployment — Deploy your agent for your team
