A template and example repository for building Claude Code plugins with MCP (Model Context Protocol) servers. Includes scaffold, pro templates, and a complete task-tracker example plugin.
Six tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. All tools have descriptions (10-50 chars), but most are generic and lack actionable context for LLM selection. Schemas present but incomplete: many parameters lack descriptions, enums are missing for constrained inputs (e.g., priority, status), and output schemas are not documented. No error handling guidance, no recovery paths. Naming follows verb_noun convention (task_add, task_list, task_complete, my_config_read, my_config_update, my_items_list) but lacks specificity in some cases (my_config_*, my_items_list are vague about what config/items mean). No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Composition is reasonable, tools are single-purpose, but parameter relationships are undocumented.
Read the current plugin configuration
Update plugin configuration values
List all items, optionally filtered by status
Add a new task to the tracker
Mark a task as completed by its ID
List all tasks, optionally filtered by status or priority
Output schemas not documented. LLMs cannot infer what fields task_add, task_list, task_complete, my_config_read, my_config_update, and my_items_list return. This forces agents to guess at response structure and breaks downstream tool chaining.
Constrained inputs lack enums. 'priority' accepts 'low, medium, high, critical' but is defined as free-form string, not enum. 'status' similarly lacks enum constraint. LLMs will hallucinate invalid values.
Parameter descriptions missing or incomplete. 'updates' in my_config_update has no description of what keys are valid or what values are expected. 'status' in my_items_list lacks description of valid values.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | 2026-07-28+ | v2 |
Vague tool names. 'my_config_read', 'my_config_update', 'my_items_list' do not clearly convey what config or items are. LLMs cannot distinguish these from similar tools without reading full descriptions. Rename to 'read_plugin_config', 'update_plugin_config', 'list_plugin_items'.
No error handling guidance. Tools do not document what errors can occur, whether they are retryable, or what the LLM should do next. E.g., task_complete does not explain what happens if task_id is invalid.