Model Context Protocol server for Sempods pod operations, providing tools for reading and writing semantic data across connected pods
Sempods MCP demonstrates solid tool design with consistent naming (verb_noun pattern), comprehensive parameter schemas, and clear descriptions. All 14 tools have descriptions (avg ~80 chars) and input schemas with typed parameters. However, output schemas are not documented in the source, parameter descriptions lack constraint details (ranges, enums, formats), and error handling guidance is absent. No tool annotations (readOnlyHint/destructiveHint) despite clear risk classifications. Composition is strong, tools chain well via IRIs, but descriptions could be more LLM-optimized per the 50-200 char baseline.
Add a value to a property (slot) of a resource
Remove all values from a property (slot) of a resource
Create a new resource in a context
Delete a resource from a context
Search for resources by text across readable contexts
Retrieve property values for a subject from readable contexts
Retrieve a resource by IRI from readable contexts
Output schemas not documented. LLMs cannot plan downstream calls or extract chaining IDs (e.g., resource_iri from create_resource). Source shows only input schemas; return types are inferred from implementation.
Parameter descriptions lack constraint details. 'limit' in find() has no min/max (1-100 stated in description but not in schema). 'query' in sparql_select has no format guidance. Enums missing for optional filters.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
List all readable contexts in this pod
List all connected pods (hosted service only)
Remove a specific value from a property (slot) of a resource
Replace all values of a property (slot) of a resource
Execute a SPARQL CONSTRUCT or DESCRIBE query against readable contexts
Execute a SPARQL SELECT or ASK query against readable contexts
Update an existing resource in a context using JSON Merge Patch
No tool annotations despite clear risk classifications. delete_resource, update_resource, and create_resource marked WRITE/DESTRUCTIVE but lack destructiveHint or idempotentHint annotations in schema. Agents cannot infer safety properties.
Error handling guidance absent. No recovery hints in descriptions (e.g., 'If resource not found, try find() first'). No indication of retryable vs. fatal errors. Agents cannot self-correct on failures.
Pagination not documented for find(). Returns up to 100 results but no next_cursor or offset mechanism visible. Large result sets could exhaust context window.