Loading ToolBench

Reading the current corpus…

vibeguard MCP server · F · ToolBench

vibeguard

MCP Server for VibeGuard — AI tool integration for Cursor, Claude Code, Windsurf. Provides snapshot management, architecture checking, health analysis, and recovery planning.

2026-07-28+STDIO8 tools

40/100

Definition Quality42

Protocol Readiness35

SupportabilityNot measured-

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

VibeGuard MCP server has 8 tools with basic naming and descriptions, but critical gaps in schema documentation and parameter descriptions. Tool names follow verb_noun convention (vibeguard_snapshot, vibeguard_rollback, etc.), which is good. However, only 2 of 8 tools have visible input schemas in the provided code. Most parameter descriptions are minimal (1-5 words). No output schemas are documented. Error handling guidance is absent. The server lacks tool annotations (readOnlyHint, destructiveHint) despite having tools with clear risk levels (WRITE, REVERSIBLE, READ_ONLY). Descriptions are adequate (20-50 chars) but lack context on when to use each tool vs. alternatives.

Tools (8)

vibeguard_analyzeread only42/100

Analyze your project's health and get a score

vibeguard_checkread only42/100

Check your project against architecture rules

vibeguard_diffread only50/100

Show changes since a snapshot

vibeguard_list_snapshotsread only50/100

List all snapshots

vibeguard_rescueread only42/100

Generate a step-by-step recovery plan for your project

vibeguard_rollbackreversible50/100

Rollback to a previous snapshot

vibeguard_rulesread only42/100

Get information about active architecture rules

vibeguard_snapshotwrite50/100

Create a snapshot of the current project state

Issues (5)

CRITICALtool-description ↗

No input schemas visible for 4 of 8 tools (vibeguard_check, vibeguard_rules, vibeguard_analyze, vibeguard_rescue). Tools with empty input objects {} have no parameter documentation.

vibeguard_checkvibeguard_rulesvibeguard_analyzevibeguard_rescue

CRITICALresponse-shaper ↗

No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., snapshotId from list_snapshots for use in diff/rollback).

vibeguard_snapshotvibeguard_rollbackvibeguard_diffvibeguard_list_snapshotsvibeguard_checkvibeguard_rulesvibeguard_analyzevibeguard_rescue

HIGHtool-annotations ↗

Tool annotations missing. vibeguard_snapshot (WRITE), vibeguard_rollback (REVERSIBLE), and others have explicit risk levels but no destructiveHint/idempotentHint in schema. Agents cannot determine retry safety.

vibeguard_snapshotvibeguard_rollback

Recommendations

Add complete input schemas for vibeguard_check, vibeguard_rules, vibeguard_analyze, vibeguard_rescue. Even if they accept no parameters, document this explicitly with 'properties: {}' and 'required: []' in JSON Schema.
Document output schemas for all 8 tools. Example: vibeguard_list_snapshots should return {snapshots: [{id, timestamp, message, size}], total, hasMore}. This enables chaining (e.g., agent extracts snapshotId for rollback).
Add tool annotations to schema: vibeguard_snapshot and vibeguard_rollback need 'destructiveHint: true' or 'idempotentHint: true' so agents know retry behavior.
Expand parameter descriptions with constraints. vibeguard_list_snapshots 'limit' should be '1-100, default 20'. vibeguard_snapshot 'message' should clarify if required or optional.
Add error recovery guidance in tool descriptions. E.g., vibeguard_rollback: 'If snapshotId not found, call vibeguard_list_snapshots to see available IDs.' This prevents dead-end failures.

Score history

Overall score trend

First recorded score · v2 rubric

40/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	40	2026-07-28+	v2