You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
| learn_shopify_api |**Start here first** - Teaches the LLM about supported Shopify APIs and how to use this MCP server's tools to generate valid code blocks for each API. This tool makes a request to shopify.dev to get the most up-to-date instruction for how to best work with the API the user would need to use for their prompt. |
96
+
| search_docs_chunks | Search across all shopify.dev documentation to find relevant chunks matching your query. Useful for getting content from many different documentation categories, but may have incomplete context due to chunking |
97
+
| fetch_full_docs | Retrieve complete documentation for specific paths from shopify.dev. Provides full context without chunking loss, but requires knowing the exact path. Paths are provided via `learn_shopify_api`|
98
+
| introspect_admin_schema | Access and search the Shopify Admin API GraphQL schema |
99
+
100
+
## Tool Usage Guidelines
101
+
102
+
### When to use each documentation tool:
103
+
104
+
-**`learn_shopify_api`**: Always call this first when working with Shopify APIs. It provides essential context about supported APIs and generates a conversation ID for tracking usage across tool calls.
105
+
106
+
-**`search_docs_chunks`**: Use when you need to find relevant information across multiple documentation sections or when you don't know the specific path. This tool searches through chunked content which allows for better token matching within smaller content pieces, but may miss some context from individual pages.
107
+
108
+
-**`fetch_full_docs`**: Use when you need complete documentation for a specific API resource and know the exact path (e.g., `/docs/api/admin-rest/resources/product`). This provides full context without any information loss from chunking.
`This tool will take in the user prompt, search shopify.dev, and return relevant documentation and code examples that will help answer the user's question.`,
162
162
withConversationId({
163
163
prompt: z.string().describe("The search query for Shopify documentation"),
`Use this tool to retrieve a list of documents from shopify.dev.`,
184
+
"fetch_full_docs",
185
+
`Use this tool to retrieve a list of full documentation pages from shopify.dev.`,
186
186
withConversationId({
187
187
paths: z
188
188
.array(z.string())
189
189
.describe(
190
-
`The paths to the documents to read, i.e. ["/docs/api/app-home", "/docs/api/functions"]. Paths should be relative to the root of the developer documentation site.`,
190
+
`The paths to the full documentation pages to read, i.e. ["/docs/api/app-home", "/docs/api/functions"]. Paths should be relative to the root of the developer documentation site.`,
// This tool is the entrypoint for our MCP server. It has the following responsibilities:
289
289
290
290
// 1. It teaches the LLM what Shopify APIs are supported with this MCP server. This is done by making a remote request for the latest up-to-date context of each API.
"conversationId. If not provided, a new conversation ID will be generated for this conversation. This conversationId should be passed to all subsequent tool calls within the same chat session.",
317
+
"Optional existing conversation UUID. If not provided, a new conversation ID will be generated for this conversation. This conversationId should be passed to all subsequent tool calls within the same chat session.",
0 commit comments