Loading ToolBench

Reading the current corpus…

query_blogger_mcp_server MCP server · D · ToolBench

← Browse

query_blogger_mcp_server

github.com/karthiks/query_blogger_mcp_server

Model Context Protocol (MCP) server for read-only, domain-specific Blogger API access.

<=2025-11-25STREAMABLE_HTTP4 tools

55/100

Definition Quality66

Protocol Readiness72

Supportability26

Spec posture

Inferred effective spec <=2025-11-25·current baseline 2026-07-28

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

Server defines 4 blog query tools with HTTP transport via FastMCP/Uvicorn. All tools have descriptions and input schemas visible in code. Naming follows verb_noun convention (get_, search_, list_). However, schemas lack completeness in parameter descriptions, output schemas are not explicitly documented, and error handling is minimal. Tool annotations are present (readOnlyHint, openWorldHint) which is good. The server implements domain-based access control, which is a security positive. Overall: functional definitions with moderate gaps typical of C-tier community servers.

Tools (4)

get_blog_info_by_urlread onlyauthsource verified66/100

Retrieves public information about a Blogger blog given its URL. ONLY works for allowed, pre-configured domains.

get_recent_postsread onlyauthsource verified65/100

Fetches the most recent public blog posts for a specified blog URL. ONLY works for allowed, pre-configured domains.

list_recent_postsread onlyauthsource verified64/100

Fetches the most recent public blog posts (Titles only) for a specified blog. ONLY works for allowed, pre-configured domains.

search_postsread onlyauthsource verified67/100

Searches for blog posts in a specified blog by query terms. ONLY works for allowed, pre-configured domains.

Issues (5)

HIGHtool, response-shaper ↗

Output schemas not documented. Tools return dicts but LLMs cannot see what fields to expect in responses. This forces LLMs to guess at downstream field names and breaks tool chaining.

get_blog_info_by_urlget_recent_postslist_recent_postssearch_posts

HIGHconstrained-input, tool-description ↗

Parameter descriptions are minimal and do not explain format/constraints. 'blog_url' lacks guidance on expected format (http vs https, domain suffix). 'query_terms' has no length limit or character restrictions documented. 'num_posts' lacks minimum/maximum bounds.

get_blog_info_by_urlget_recent_postslist_recent_postssearch_posts

MEDIUMrecovery-guide, error-classification ↗

Error handling is present but not actionable. The code returns dict errors like {'error': 'Access denied...'} and {'error': 'Failed to retrieve...'} but does not provide recovery guidance (e.g., 'Try a different domain' or 'Check blog URL format').

get_blog_info_by_url

Recommendations

Document the return type for each tool. Example for get_blog_info_by_url: 'Returns {blog_id: str, blog_title: str, description: str, published: str (ISO 8601), url: str} or {error: str, requested_url: str} on failure.'
Add constraints to parameter descriptions. For blog_url: 'Full HTTPS URL of the blog (e.g. https://example.blogspot.com). Must be from an allowed domain.'
Add bounds to numeric parameters. For num_posts: 'Maximum number of posts to retrieve. Must be between 1 and 50 (default: 3).'
Enhance error messages with recovery guidance. Instead of 'Access denied: This tool can only query blogs from pre-approved domains', return 'Domain not allowed. Contact admin with blog URL or use a blog from the approved list (check settings.ALLOWED_DOMAINS).'
Add a helper tool or list endpoint to discover allowed domains. Current implementation has no way for LLMs to know which blogs are queryable.
Document the response schema in the tool docstring, including error cases. Use structured output format (JSON schema comment) so LLMs know what fields to expect.
Consider adding pagination support. If blogs have many posts, returning all by default will bloat responses. Add a 'next_cursor' or 'offset' field to enable continuation.

Spec posture evidence

Inferred effective spec: <=2025-11-25.

Relies on Logging (deprecated) - log to stderr or use OpenTelemetry

Score history

Overall score trend

↑ 55 points across a rubric change (v1 → v2)

55/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-22	D	55	<=2025-11-25	v2
2026-03-09	F	0	-	v1