MCP server for Worktree Manager - provides tools for managing git worktrees, workspaces, and projects
Worktree Manager MCP has 13 tools with basic naming and schema structure, but significant gaps in descriptions and parameter documentation. All tools have names starting with action verbs (list_, create_, delete_, etc.), which is good. However, most tool descriptions are extremely brief (10-30 chars), falling well below the 50-200 char baseline for LLM-optimized descriptions. Parameter descriptions exist but are minimal ('Name of the worktree'). Input schemas are present with type definitions, but lack constraints (enums, ranges, patterns). No output schemas are documented. Error handling is absent, no recovery guidance, no categorization of error types. The tool set is well-composed (single responsibility per tool, good naming distinctions), but the descriptions are too terse to guide LLM selection effectively.
Add a project to a worktree
Archive a worktree
Create a new git worktree
Delete a worktree
Get the status of a project in a worktree
Get the status of a worktree
List all projects in a worktree
List all worktrees in the current workspace
Lock a worktree to prevent modifications
Tool descriptions are critically under-specified (10-30 chars). LLMs cannot determine when to select a tool vs. similar alternatives. Baseline is 50-200 chars with context on WHAT, WHEN, and WHY.
No output schemas documented. LLMs cannot plan downstream tool calls or extract required fields (e.g., worktree_id for chaining). Every tool must document its return structure.
No error handling or recovery guidance. Tools do not indicate what errors are retryable, user-fixable, or fatal. No actionable error messages (e.g., 'Worktree not found. Available: [list]').
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | 2026-07-28+ | v2 |
Remove a project from a worktree
Restore an archived worktree
Switch to a different worktree
Unlock a worktree
Destructive operations (delete_worktree, remove_project) lack confirmation or dry-run support. Agents can accidentally destroy data without a safety gate.
Parameter descriptions are minimal ('Name of the worktree'). No constraints documented (e.g., name length, allowed characters, format). LLMs cannot validate input before calling.