Connect to your Salla store to manage products, orders, customers, and analytics through AI assistants.
The Salla MCP server has 18 tools with generally present descriptions and schemas, but with significant inconsistencies in quality. Strong points: Tools 1-3 (orders) and 4 (products list) have detailed, LLM-friendly descriptions with explicit guidance on when to use them and emoji-enhanced readability. Tool 1 (salla-orders-list) exemplifies good practice with 16 parameters, each annotated with emoji prefixes and concrete usage scenarios. Weak points: Tools 5 (product details), 8-9 (customers), 10-12 (store/categories) have minimal descriptions (5-60 chars), violating the 50-200 char baseline for LLM optimization. Tools 14-18 (reports) use a placeholder 'random_string' parameter to work around no-parameter endpoints, which is a design anti-pattern. Parameter descriptions vary widely in specificity, Tool 1 explains each filter's purpose and valid values (e.g., 'SEARCH:', 'FILTER:', 'DATE FILTER:'), while Tools 5-12 use generic 1-2 word descriptions like 'The ID of the product to retrieve' with no guidance on format or constraints. Output schemas are not documented in the tool definitions, making it impossible for agents to plan downstream calls. Error handling is absent, no guidance on retryability, user-fixable errors, or recovery paths. Security: credentials are properly injected server-side (evident from OAuth handler), not exposed as parameters. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite several read-only tools and two write operations (salla-order-status-update, salla-product-create, salla-product-update).
Get a report of abandoned carts from your Salla store.
Get a list of brands from your Salla store.
Get a list of product categories from your Salla store.
Get detailed information about a specific product category.
Get detailed information about a specific customer by their ID.
Get a list of customers from your Salla store. You can search by name or email and paginate through results.
Get general statistics report from your Salla store.
Placeholder 'random_string' parameter used as workaround for no-parameter endpoints in tools 14-18 (reports). This violates the principle of explicit, minimal required parameters and confuses agents about what they're actually controlling.
Output schemas are not documented in any tool definition. LLMs cannot plan downstream tool calls or extract required IDs without knowing what fields to expect. For example, salla-orders-list should document that it returns order_id, customer_id, status, total, created_at, etc., so agents can reference these in subsequent calls.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 55 | 2026-07-28+ | v2 |
Get hourly visitors report from your Salla store.
Get the latest orders report from your Salla store.
🔍 Get comprehensive details about a specific order. Use this when user asks for full order information, customer details, items purchased, shipping info, etc. Usually called after getting order ID from orders list.
⚡ Update the status of a specific order in your Salla store. Use when user asks to mark orders as shipped, delivered, cancelled, etc. Important: Only use valid Salla status values.
📋 Get a list of orders from your Salla store with powerful filtering options. Perfect for finding orders by customer, date range, status, or other criteria. Supports all common AI chat scenarios like 'show me orders from John', 'recent orders', 'pending orders', etc.
Create a new product in your Salla store.
Get detailed information about a specific product by its ID.
Update an existing product in your Salla store.
📦 Get a list of products from your Salla store with advanced filtering options. Perfect for inventory management, product searches, and category browsing. Supports all common AI chat scenarios like 'show me hidden products', 'products on sale', 'search for iPhone', etc.
Get information about your Salla store.
Get a summary report from your Salla store. Only the 'monthly' period is supported.
Minimal descriptions for tools 5, 8-12, 14-18. Tool 5 (salla-product-details) has description 'Get detailed information about a specific product by its ID.' (71 chars), below the 50-200 char baseline. Tool 10 (salla-store-info) has NO description at all. Agents cannot distinguish when to use product-details vs products-list, or when store-info is appropriate.
No error handling guidance in any tool. Agents do not know if errors are retryable, user-fixable, or fatal. E.g., salla-order-status-update fails silently if given an invalid status string, no guidance on valid values or how to recover.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Tools like salla-orders-list are read-only but unmarked. Tools salla-order-status-update, salla-product-create, salla-product-update are destructive/state-changing but unmarked. Agents cannot distinguish safe tools from risky ones.
Parameter descriptions lack format constraints, enums, and range validation. E.g., 'from_date' and 'to_date' accept YYYY-MM-DD but no validation is shown; agents could pass invalid dates. 'page' accepts 'number' with no min/max, agents could pass 0 or negative values. Enum constraints should be declared in schema, not just mentioned in descriptions.
Tool composition: salla-orders-list returns up to 15 items per page (default) with no mention of result limits in description. If a user asks 'show me all orders', the agent may loop through all pages without context limits. Description should state: 'Returns up to 15 orders per page (max 50 per_page). Use pagination to fetch all results in batches.'
salla-order-status-update accepts 'status' as free-form string with valid values mentioned in description ('pending', 'under_review', 'in_progress', 'shipped', 'delivered', 'cancelled') but NOT declared as enum in schema. LLMs may hallucinate invalid statuses like 'processing' or 'shipped_pending'. Use enum constraint in schema definition.