Model Context Protocol server for Redis - provides tools for reading and writing to Redis databases
Redis MCP server has 30 tools with consistent naming (verb_noun pattern: get_, set_, push_, pop_, etc.) and descriptions present for all tools. However, parameter descriptions are often minimal (5-15 chars), and output schemas are not documented. Error handling returns generic error strings rather than actionable recovery guidance. Tool composition is sound (each tool does one thing), but parameter validation rules and constraints are underspecified in descriptions. No tool annotations (readOnlyHint/destructiveHint) despite clear risk classifications in metadata.
Get a list of connected clients to the Redis server.
Get the number of keys stored in the Redis database
Execute a Lua script on the Redis server.
Execute a Lua script by its SHA1 digest.
Retrieve a vector from a Redis hash and convert it back from binary blob.
Delete a field from a Redis hash.
Check if a field exists in a Redis hash.
Output schemas not documented. Tools return raw Redis responses (dicts, lists, strings) without specifying field names, types, or structure. LLMs cannot plan downstream tool calls or extract specific fields reliably.
Parameter descriptions are minimal (5-15 chars) and lack validation constraints. E.g., 'section' in info() says 'The section of the info command (default, memory, cpu, etc.)' but does not list valid enum values or explain what each returns. LLMs cannot validate inputs before calling.
Error handling returns generic error strings ('Error getting database size: ...') without recovery guidance. LLMs cannot determine if errors are retryable, user-fixable, or fatal. No actionable next steps provided.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 67 | <=2025-11-25 | v2 |
Get the value of a field in a Redis hash.
Get all fields and values from a Redis hash.
Set a field in a hash stored at key with an optional expiration time.
Get Redis server information and statistics.
Append a value to a JSON array at a specific path.
Get the length of a JSON array at a specific path.
Pop (remove and return) an element from a JSON array at a specific index.
Delete a JSON value from Redis at a given path.
Retrieve a JSON value from Redis at a given path.
Increment a number value in a JSON document by a specific amount.
Get the keys of a JSON object at a specific path.
Set a JSON value in Redis at a given path with an optional expiration time.
Toggle a boolean value in a JSON document.
Get the length of a Redis list.
Remove and return the first element from a Redis list.
Push one or more values onto the left of a Redis list and optionally set an expiration time.
Get elements from a Redis list within a specific range.
Remove and return the last element from a Redis list.
Push one or more values onto the right of a Redis list and optionally set an expiration time.
Check if a script exists in the script cache.
Flush the Redis script cache.
Load a Lua script into the Redis script cache.
Store a vector as a field in a Redis hash.
Tool annotations missing. Despite clear risk classifications (READ_ONLY, WRITE, DESTRUCTIVE), tools lack readOnlyHint, destructiveHint, or idempotentHint annotations in schema. Clients cannot infer safety properties from tool definitions.
Lua script tools (eval_script, evalsha_script) accept arbitrary script code and arguments but do not document what keys/args the script expects or what it returns. LLMs cannot reason about script safety or output structure.