MCP server for SET (ShipExactlyThis) - exposes worktree and Ralph status to AI agents
Three tools with minimal documentation. All have descriptions (10-50 chars), but lack parameter descriptions and output schemas. Input schemas are present but sparse. No error handling guidance, no parameter constraints, no composition hints. Naming is acceptable (verb_noun pattern), but descriptions are too brief to guide LLM selection. No evidence of idempotency, pagination, or chaining support.
Get Ralph loop status for a worktree. Args: change_id: Optional change ID to check. If not provided, shows all active loops. Returns status including: running/stopped, iteration count, last activity.
Get tasks for a worktree. Args: worktree_path: Path to the worktree directory. Returns the tasks.md content showing implementation progress.
List all git worktrees across all projects. Shows worktree path, branch, and status for each active worktree.
No output schemas documented. LLMs cannot infer what fields to expect or plan downstream tool calls. All three tools lack return type documentation.
Parameter descriptions missing or incomplete. 'change_id' and 'worktree_path' lack detail on format, constraints, or examples. LLMs cannot validate inputs.
Tool descriptions too brief (10-50 chars). Baseline is 194 chars. Descriptions lack WHEN to use, WHAT it returns, and prerequisites. Insufficient for LLM selection.
No error handling guidance. Tools do not document what happens on failure (missing worktree, invalid change_id, file not found). LLMs have no recovery path.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | 2026-07-28+ | v2 |
No pagination or result limits documented. list_worktrees may return unbounded results, risking context window exhaustion.