Loading ToolBench
MCP (Model Context Protocol) server for Redd-Archiver API - exposes 30+ archive endpoints (Reddit, Voat, Ruqqus) as AI-accessible tools with MCP resources
This server provides only 2 tools, both of which are READ_ONLY guidance/reference tools rather than functional API tools. No actual data-manipulation or query tools are visible in the provided source code. The tool descriptions are minimal, parameter schemas are absent, and output structures are undocumented. The server is intended to expose 30+ archive endpoints per the README, but only 2 guidance tools appear in the provided code snippet. Without seeing the actual tool definitions, schemas, and descriptions for the claimed 30+ tools, a comprehensive quality assessment is impossible. Based on what IS visible: both tools lack input parameters (they are read-only guides), lack documented output schemas, and have minimal descriptions (under 200 chars). This falls well below production quality standards.
Most recent 25 posts across all subreddits.
Search operator documentation and examples. Returns help text explaining supported search operators.
Current archive statistics (posts, comments, users, subreddits).
List of all subreddits in the archive with post counts.
Top 25 posts by score across all subreddits.
How to build effective search queries with operators. Use this to learn search syntax before calling Full_text_search.
Best practices for avoiding token overflow when using Redd-Archiver MCP tools. READ THIS BEFORE using Full_text_search, List_posts, or List_comments tools.
No input parameter schemas visible. Both tools lack documented input parameters, types, and constraints. Unable to verify tool definitions are properly registered with JSON Schema.
No output schemas documented. Tools do not declare what data structure they return, preventing LLMs from planning downstream operations or extracting structured results.
Both tools are READ_ONLY reference/guidance tools, not functional data tools. The README claims 30+ archive endpoints are exposed, but the provided source code shows only these 2 guidance tools. Core functional tools (Full_text_search, List_posts, List_comments, etc. mentioned in token-safety-guide description) are not visible in the provided code snippet.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 18 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 40 | - | v1 |
Minimal descriptions. token-safety-guide is ~125 chars, search-query-builder is ~95 chars. While acceptable, descriptions do not explain WHEN to use each tool vs. the other, or what output format to expect. Descriptions should be 50-200 chars with explicit action statements.
No error handling or recovery guidance. Tools lack description of failure modes, what could go wrong, or how to recover if referenced tools (Full_text_search, List_posts, List_comments) fail.
Naming convention concerns. 'token-safety-guide' and 'search-query-builder' are hyphenated nouns, not verb-noun pairs (get_*, list_*, create_*). They read as resource names rather than callable actions. LLMs expect action verbs to understand tool intent.