Skip to content

contentstack/contentstack-vibe-docs

Repository files navigation

Contentstack Vibe Docs

An Agent Skill that gives AI coding agents comprehensive Contentstack CMS knowledge. Works with Claude Code, Cursor, GitHub Copilot, VS Code, Gemini CLI, Roo Code, and 25+ other tools.

AI Agents: Read skills/contentstack-vibe-docs/SKILL.md. It contains routing rules that direct you to the 1-3 files you need. Do not read all files — this skill contains ~13,500 lines across 34 reference documents.


Install

Universal (all agents)

Install with the skills CLI:

npx skills add contentstack/contentstack-vibe-docs

This auto-detects your installed agents (Claude Code, Cursor, Copilot, Windsurf, Cline, etc.) and installs the skill for all of them.

Install for specific agents only:

npx skills add contentstack/contentstack-vibe-docs -a claude-code
npx skills add contentstack/contentstack-vibe-docs -a cursor
npx skills add contentstack/contentstack-vibe-docs -a github-copilot

Install globally (available in all your projects):

npx skills add contentstack/contentstack-vibe-docs -g

Claude Code

Install as a personal skill (all projects):

claude skill add contentstack/contentstack-vibe-docs

Or use the skills CLI:

npx skills add contentstack/contentstack-vibe-docs -a claude-code -g

Once installed, Claude Code discovers the skill automatically when you work on Contentstack tasks. You can also invoke it directly:

/contentstack-vibe-docs fetch blog posts with author references
/contentstack-vibe-docs set up live preview in Next.js
/contentstack-vibe-docs optimize images for responsive layouts

Gemini CLI

gemini skills install https://github.com/contentstack/contentstack-vibe-docs.git

Install to workspace scope:

gemini skills install https://github.com/contentstack/contentstack-vibe-docs.git --scope workspace

Manage installed skills

npx skills list              # List all installed skills
npx skills check             # Check for updates
npx skills update            # Update all skills

Claude Code Plugin

This repo is also a Claude Code plugin. Install it directly:

claude plugin install https://github.com/contentstack/contentstack-vibe-docs

Or test locally during development:

claude --plugin-dir ./contentstack-vibe-docs

The plugin includes:

  • Skills — the full Contentstack documentation skill with routing and references

Cursor Plugin

This repo is also a Cursor plugin. Install it directly in Cursor:

  1. Open Cursor Settings > Plugins
  2. Add this repository URL: https://github.com/contentstack/contentstack-vibe-docs

The plugin includes:

  • Skills — the full Contentstack documentation skill with routing and references
  • Rulesrules/contentstack.mdc with routing table, decision helpers, security guardrails, and quick patterns

Cursor (manual install)

If you prefer not to use the plugin marketplace, copy the .mdc rule file directly into your project:

git clone --depth 1 https://github.com/contentstack/contentstack-vibe-docs.git /tmp/cs-skills && \
  mkdir -p .cursor/rules && \
  cp /tmp/cs-skills/cursor/rules/*.mdc .cursor/rules/ && \
  rm -rf /tmp/cs-skills

Cursor auto-discovers .cursor/rules/*.mdc on project open.

Codex CLI

Codex auto-discovers AGENTS.md in your project root. Copy the prebuilt codex/ directory in:

git clone --depth 1 https://github.com/contentstack/contentstack-vibe-docs.git /tmp/cs-skills && \
  cp -r /tmp/cs-skills/codex . && \
  rm -rf /tmp/cs-skills

This places codex/AGENTS.md and codex/skills/contentstack-vibe-docs/ in your project. Codex reads AGENTS.md, which routes it to the skill.

Any other agent (manual)

Every major agent reads at least one of: CLAUDE.md, AGENTS.md, skills/*/SKILL.md, or .cursor/rules/*.mdc. Point your agent at the repo root or copy skills/contentstack-vibe-docs/ into your project — the routing table in SKILL.md does the rest.


How It Works

This skill uses progressive disclosure to keep your agent's context window clean:

  1. Discovery (~100 tokens) — The agent loads the skill name and description. Just enough to know "this is relevant for Contentstack tasks."
  2. Activation (~3,500 tokens) — When a Contentstack task is detected, the agent reads SKILL.md with its routing table, decision helpers, and inline quick-start patterns.
  3. Execution (on-demand) — The routing table directs the agent to read only the 1-3 specific reference files it needs. It never loads all ~13,500 lines.
User: "Add live preview to my Next.js app"
  → Agent loads skills/contentstack-vibe-docs/SKILL.md (routing table)
  → Reads skills/contentstack-vibe-docs/references/live-preview/concepts.md
  → Reads skills/contentstack-vibe-docs/references/live-preview/ssr-mode.md
  → Reads skills/contentstack-vibe-docs/references/frameworks/nextjs.md
  → Implements with copy-paste ready code

What's Covered

Topic Reference File
Quick code patterns QUICK_REFERENCE.md
CMS fundamentals base-concepts.md
Data modeling best practices data-modeling-best-practices.md
Region configuration regions.md
REST Content Delivery API rest-api.md
GraphQL Content Delivery API graphql-api.md
Content Management API (CRUD) content-management-api.md
Image Delivery API (transforms) image-delivery-api.md
TypeScript Delivery SDK delivery-sdk.md
Live Preview (concepts) concepts.md
Live Preview (CSR mode) csr-mode.md
Live Preview (SSR mode) ssr-mode.md
Next.js patterns nextjs.md
Nuxt patterns nuxt.md
Gatsby patterns gatsby.md
OAuth with Auth.js oauth.md
CLI plugin development cli-plugins/overview.md
Developer Hub apps devhub-apps.md
Real-world code examples practical-examples.md
Package version compatibility VERSIONS.md

All regions supported: US, EU, AU, Azure NA/EU, GCP NA/EU.


Security & Guardrails

This skill includes built-in security measures and red flags that instruct agents to handle credentials safely:

  • No secret leaking — Agents are explicitly told to never ask for, output, log, or hardcode API keys, tokens, or secrets. All code uses process.env.* references.
  • Credential hygiene — If a developer accidentally pastes a real token in the conversation, the agent is instructed to warn them and suggest they rotate it.
  • Frontend protection — Management Tokens are flagged as server-side only. Agents are told to never expose them in frontend code.
  • .env safety — Agents are instructed to never commit .env files or credentials to version control.
  • Red flags checklist — A dedicated "Red Flags" section in SKILL.md lists specific anti-patterns agents must avoid, from hardcoding secrets to mixing SDK patterns incorrectly.
  • Questions before code — Agents are guided to ask about region, framework, and environment setup before writing any code, ensuring the right patterns are used from the start.

Documentation Structure

This repo follows the Open Plugins standard and the agentskills.io skill-directory convention. The skill itself is canonical — everything else is either a root-level pointer or a generated per-harness copy.

CLAUDE.md                    # Root pointer for Claude Code (project context)
AGENTS.md                    # Root pointer for agentskills.io / Codex / generic agents
README.md                    # This file
gemini-extension.json        # Manifest for `gemini skills install`

.plugin/plugin.json          # Open Plugins manifest
.claude-plugin/plugin.json   # Claude Code plugin manifest
.cursor-plugin/plugin.json   # Cursor plugin manifest

skills/
└── contentstack-vibe-docs/  # ★ Source of truth — the skill itself
    ├── SKILL.md             # Routing table + decision helpers
    └── references/          # 34 reference files (see tree below)

rules/
└── contentstack.mdc         # Source — Cursor rule (points at SKILL.md)

codex/                       # Generated — copy into your project for Codex CLI
├── AGENTS.md                # (hand-written, not generated)
└── skills/contentstack-vibe-docs/
    ├── SKILL.md
    └── references/

cursor/
└── rules/contentstack.mdc   # Generated — for manual `.cursor/rules/` install

scripts/
└── build.sh                 # Regenerates codex/ and cursor/ from the canonical source

skills/contentstack-vibe-docs/references/
├── QUICK_REFERENCE.md          # Condensed patterns for quick lookup
├── VERSIONS.md                 # Package version compatibility
├── concepts/
│   ├── base-concepts.md        # CMS fundamentals
│   ├── data-modeling-best-practices.md  # Schema design + taxonomy CDA operators
│   ├── localization.md         # Master language, fallback chains, non-localizable fields
│   └── regions.md              # Region configuration
├── api/
│   ├── rest-api.md             # REST Content Delivery API
│   ├── graphql-api.md          # GraphQL Content Delivery API
│   ├── content-management-api.md  # CMA (CRUD, modular block schema, headers)
│   └── image-delivery-api.md   # Image transforms + asset organization + limits
├── sdk/
│   └── delivery-sdk.md         # TypeScript SDK guide (region-aware preview)
├── live-preview/
│   ├── concepts.md             # Live Preview overview
│   ├── csr-mode.md             # ssr:false — postMessage updates
│   ├── ssr-mode.md             # ssr:true — per-request factory pattern
│   ├── visual-builder.md       # Edit tags, addEditableTags, VB_EmptyBlockParentClass
│   └── debugging.md            # Symptom-based diagnostic guide
├── authentication/
│   └── oauth.md                # OAuth login with Auth.js v5 (Next.js)
├── security/
│   ├── tokens-authentication.md  # Delivery/Preview/Management/Authtoken/OAuth decision tree
│   └── roles-permissions.md    # Built-in roles, custom roles, teams
├── workflows/
│   ├── webhooks.md             # Event channels, signature verification, release storms
│   ├── releases.md             # Atomic coordinated content deployment
│   ├── content-workflows.md    # Workflow stages, publish rules, approval gates
│   ├── branches-aliases.md     # Zero-downtime deployment via aliases
│   └── environments-publishing.md  # Publishing lifecycle, Sync API, rate limits
├── personalization/
│   └── variants-and-personalize.md  # Variants vs separate entries + Personalize SDK
├── frameworks/
│   ├── nextjs.md               # Next.js patterns + Draft Mode + revalidateTag
│   ├── nuxt.md                 # Nuxt 4 patterns
│   └── gatsby.md               # Gatsby patterns
├── extensions/
│   ├── cli-plugins/            # CLI plugin development (overview + commands + publishing)
│   │   ├── overview.md
│   │   ├── commands.md
│   │   └── publishing.md
│   ├── devhub-apps.md          # Developer Hub apps (App SDK, UI locations, API proxy)
│   └── launch.md               # Contentstack Launch deployments + env sync
└── examples/
    └── practical-examples.md   # Real-world code patterns

Maintainer note

The single source of truth is skills/contentstack-vibe-docs/ (SKILL.md + references/) and rules/contentstack.mdc. Everything under codex/skills/ and cursor/rules/ is generated — do not edit those directly. After editing a source file, run:

bash scripts/build.sh

Generated files are committed so GitHub / zip-download installs work without running the script locally.


Support


License

This documentation is provided as-is for use with AI coding assistants.

About

No description, website, or topics provided.

Resources

License

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages