One-click Netlify deployment of Corsen Context with static HTML and a serverless MCP endpoint.
Four tools with basic verb-noun naming (search_site, get_page_content, list_content, get_sitemap). All have short descriptions (10 - 50 chars) and input schemas with type definitions. However, descriptions lack depth, they do not explain WHEN to use each tool, dependencies, or error recovery. Parameter descriptions are minimal (e.g., 'Search query, 1 - 500 code points' lacks context on what 'search' means in this domain). Output schemas are not documented. No error handling guidance. Schemas are present but sparse; list_content lacks enum for 'type' parameter, inviting invalid values. Overall: functional but below production baseline (194 chars avg description; 100% of A+ tools document outputs).
One public URL as clean Markdown
Bounded structured overview
One allowed content type, paged
Matching public URLs and snippets
Descriptions are too brief (10 - 50 chars) and lack context. 'Matching public URLs and snippets' does not explain when to call search_site vs list_content, what 'matching' means, or what the response structure is.
Output schemas are not documented. LLMs cannot plan downstream calls or extract fields without knowing what search_site returns (e.g., does it return URLs, snippets, both? What fields?).
list_content 'type' parameter lacks enum constraint. Free-form string invites hallucinated content types. Should enumerate valid types (e.g., 'blog', 'docs', 'product') to guide LLM selection.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | 2026-07-28+ | v2 |
No error handling guidance. If search_site returns no results, or get_page_content hits a 404, what should the LLM do next? No recovery hints provided.
Parameter descriptions are minimal. 'URI, 1 - 2000 code points, same-site only' lacks explanation of what 'same-site' means or why it matters. Should clarify: 'Must be a relative path or absolute URL on the same domain; external URLs are rejected.'