PI-Desktop exposes 3 plugin development tools with minimal documentation. All tools have descriptions (10-20 chars each), but they are vague and lack actionable context. Input schemas are present with basic type info, but parameter descriptions are generic ('Target directory path') without constraints, examples, or guidance on valid values. No output schemas are documented. Error handling is absent, no guidance on recovery, retryability, or what happens on failure. Tool names lack clear verb-noun structure (PluginScaffold, PluginCheck, PluginPack are noun-heavy). No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk profiles (WRITE vs READ_ONLY). Composition is reasonable, each tool has one responsibility, but the interface is not LLM-optimized.

• Expand tool descriptions to 50-200 characters. Example: 'PluginScaffold → create_plugin: Generate a new plugin from a template (available: react, vue, vanilla). Returns plugin_id and directory path. Use this to bootstrap a new plugin project.'
• Document output schemas for all tools. Specify what fields are returned (e.g., plugin_id, path, status, errors) and their types. This enables LLMs to chain tools and extract results.
• Add enum constraint to 'template' parameter with valid template names. Replace 'must be one of the valid template names' with explicit list: ['react', 'vue', 'vanilla'] (or actual values).
• Add error handling guidance to descriptions. E.g., 'Returns error if directory exists (use force=true to overwrite) or template is invalid. Call list_templates to see available options.'
• Rename tools to verb-noun form: PluginScaffold → create_plugin, PluginCheck → validate_plugin, PluginPack → package_plugin. This clarifies intent and improves LLM tool selection.