Loading ToolBench
An MCP server for folder icon management and naming automation on macOS. Provides tools for scanning folders, applying custom icons, managing icon rules, learning folder name preferences, and managing automation hooks.
Folico has 21 well-named tools with consistent verb_noun patterns (folico_scan_folder, folico_apply_icons, etc.). All tools have descriptions (avg ~120 chars, within baseline 34-392 range). Input schemas are present with typed parameters and descriptions. However, output schemas are NOT documented, critical gap for LLM planning. Many parameter descriptions lack format/constraint details (e.g., 'path' doesn't specify absolute vs relative, glob patterns for exclusions lack examples). Error handling is minimal, no recovery guidance or actionable error messages visible. Confirmation patterns exist (confirmApply, confirmRestore) but lack dry-run alternatives. No tool annotations (readOnlyHint, destructiveHint) despite clear risk levels in metadata.
Applies custom folder icons to selected folders based on the scan results. Requires confirmation via confirmApply parameter.
Retrieves the current Folico application settings and configuration.
Lists all folder exclusion patterns.
Lists the history of icon changes applied to folders.
Lists all configured automation hooks.
Lists all configured folder icon rules.
Records a user's icon preference for a folder, used for learning icon matching rules.
Output schemas not documented. LLMs cannot plan downstream calls or extract required fields (e.g., what does folico_scan_folder return? folder IDs? icon suggestions? confidence scores?). This forces agents to guess or make extra discovery calls.
Parameter descriptions lack format/constraint details. 'path' parameter doesn't specify absolute vs relative paths, glob patterns for exclusions lack examples, 'commandTemplate' for hooks lacks variable substitution docs. LLMs cannot validate inputs without explicit constraints.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 60 | <=2025-11-25 | v2 |
Records that a user accepted a suggested folder name, used for learning naming preferences.
Records that a user rejected a suggested folder name, used for learning naming preferences.
Removes a folder exclusion pattern by ID.
Removes an automation hook by ID.
Removes a folder icon rule by ID.
Restores original folder icons for previously modified folders. Requires confirmation via confirmRestore parameter.
Reviews a proposed folder naming plan and returns validation results.
Scans a folder and its subfolders to identify folders that match icon rules, returning suggestions for icon assignments.
Enables or disables an automation hook by ID.
Analyzes a folder and suggests improved folder names based on learned naming preferences and rules.
Updates Folico application settings such as auto-watch folders, notifications, learning preferences, and MCP HTTP server configuration.
Creates or updates a folder exclusion pattern to prevent scanning or icon application.
Creates or updates an automation hook that executes a command in response to specified events.
Creates or updates a folder icon rule that matches folders by keywords and assigns icons.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels in metadata. folico_apply_icons and folico_remove_hook are destructive but lack annotations to guide agent caution.
Error handling is minimal. No recovery guidance visible (e.g., 'If path not found, try folico_scan_folder first'). No actionable error messages (e.g., 'Invalid icon ID: got "icon_123", available icons: [...]'). Agents cannot self-correct on failures.
Confirmation patterns (confirmApply, confirmRestore) lack dry-run alternatives. Agents cannot preview changes before committing. No partial-failure handling for batch operations (e.g., if 3 of 10 folders fail to apply icons, what is returned?).