MCP server for accessing LINE Developers documentation and search functionality
Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. readDocumentation has a minimal 1-param schema; searchDocumentation has 3 params with enums and defaults. Both descriptions are under 100 chars and lack context on when to use each tool or what they return. No error handling guidance, no output schema documentation, and parameter descriptions are sparse. Tool names follow verb_noun convention (positive), but lack the specificity needed for LLM disambiguation (e.g., 'readDocumentation' vs 'searchDocumentation' could be clearer about scope: LINE Developers site only).
Read documentation in the LINE Developers site
Search documentation in the LINE Developers site
Output schemas not documented. Tools return structured responses but LLMs cannot infer field names, types, or whether pagination is available. readDocumentation returns {content: [{type, text}]} and searchDocumentation returns SearchResult[], neither is documented in tool definitions.
Parameter descriptions are minimal or missing context. 'url' in readDocumentation has no guidance on format or valid domain. 'searchPhrase' lacks examples or length constraints. 'locale' enum is present but no description of what happens if unsupported locale is passed.
Tool descriptions lack context on when to use each tool. Both operate on LINE Developers site but descriptions do not explain: search for discovery, read for full content? No guidance on prerequisites or dependencies.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 53 | 2026-07-28+ | v2 |
No error handling guidance. If URL is invalid, search returns no results, or locale is unsupported, tools provide no recovery hints. LLMs cannot self-correct without actionable error messages.
searchDocumentation pagination is implicit (pageNumber param exists) but no documentation of total results, next_cursor, or result limits. LLMs cannot plan multi-page queries without knowing result structure.