npx skills add ...
npx skills add contentstack/contentstack-agent-skills --skill cms-data-modeling-best-practices
npx skills add contentstack/contentstack-agent-skills --skill cms-data-modeling-best-practices
Guide developers to model content in Contentstack using the simplest reusable structure. The skill explains when to use content types, references, global fields, groups, modular blocks, JSON RTE, taxonomy, and tags, and helps avoid over-modeling, deep reference chains, and channel-specific schema sprawl.
Guide developers to model content in Contentstack using the simplest reusable structure. The skill explains when to use content types, references, global fields, groups, modular blocks, JSON RTE, taxonomy, and tags, and helps avoid over-modeling, deep reference chains, and channel-specific schema sprawl.
Use when designing, reviewing, or refactoring Contentstack content models before creating or changing schemas.
Developers need a practical way to choose the right Contentstack construct so editors can work efficiently, delivery code stays simple, and schemas stay reusable, governed, and easy to query.
Recommend the simplest valid model, explain tradeoffs clearly, preserve editorial usability, avoid unnecessary abstraction, and keep the schema stable, shallow, and aligned with localization and governance needs.
Identify the domain concept, editorial workflow, delivery channels, localization needs, reuse requirements, and query constraints before recommending changes.
Pick the simplest Contentstack construct that fits: content type, reference, global field, group, modular block, JSON RTE, taxonomy, tags, or plain field.
Use reusable, governed structures when content changes independently or appears across multiple entries. Keep parent-owned data inline.
Treat content types as API contracts. Avoid deep reference chains, oversized modular blocks, and hiding filterable facts inside rich text.
Localize only fields that need translation. Keep names clear and avoid channel-specific schema pollution.
State why the recommended option is better, what it avoids, and what maintenance or query cost it reduces.
Give a concise recommendation, compare alternatives only when useful, and include migration cautions when schema changes are implied.
Use a content type for a real domain concept with its own lifecycle. Use a reference for reusable content with independent ownership. Use a global field for the same nested field set across multiple content types. Use a group for parent-owned nested data inside one content type. Use modular blocks for page-local composition. Use JSON RTE for narrative content. Use taxonomy for governed classification. Use tags for lightweight internal labels.
Use concise, structured, instruction-oriented prose. Prefer bullets and short sections. State the recommended choice first when comparing options. Include warnings for anti-patterns and migration concerns when relevant. Do not expose secrets, API keys, or management tokens.
Read-only advisory skill. Prefer default CMS knowledge and documentation sources. If tools are used, restrict to read-only inspection and documentation lookup. Do not perform schema changes, publishing, or destructive actions.
Do not perform destructive actions. Do not delete, publish, unpublish, or modify Contentstack resources. Provide guidance only.
Treat all credentials as sensitive. Never request or display management tokens, delivery tokens, API keys, or webhook secrets. Use placeholders and environment variables only.
Use environment variables for any credentialed examples or integrations. Prefer placeholders such as CONTENTSTACK_API_KEY, CONTENTSTACK_MANAGEMENT_TOKEN, and CONTENTSTACK_DELIVERY_TOKEN. Never hardcode secrets in examples or instructions.