Roundtable
Multi-model AI debates: GPT-4o, Claude, Gemini & 200+ models discuss, then synthesize insight.
사용해야 할까요
품질 및 안전성
A
발견 사항 (14)
- LOWTool 'plan-implementation' description lacks action verbplan-implementation에서
- LOWTool 'list-models' doesn't follow camelCase/snake_caselist-models에서
- LOWTool 'list-sessions' doesn't follow camelCase/snake_caselist-sessions에서
- LOWTool 'get-session' doesn't follow camelCase/snake_caseget-session에서
- LOWTool 'get-logs' doesn't follow camelCase/snake_caseget-logs에서
- LOWTool 'check-usage' doesn't follow camelCase/snake_casecheck-usage에서
- LOWTool 'get-thread-link' doesn't follow camelCase/snake_caseget-thread-link에서
- LOWTool 'set-thread-visibility' doesn't follow camelCase/snake_caseset-thread-visibility에서
- LOWTool 'consult-council' doesn't follow camelCase/snake_caseconsult-council에서
- LOWTool 'design-architecture' doesn't follow camelCase/snake_casedesign-architecture에서
도구 정의와 프로토콜 준수에 대한 자동 분석을 기반으로 합니다.
컨텍스트 비용
~4,423토큰 (도구 정의)
~3.6 KB일반적인 응답 크기
상당한 주의 영향 (128k 컨텍스트의 3.46%)
이는 서버의 도구가 모델의 컨텍스트에 로드될 때마다 소비되는 대략적인 토큰 수입니다. 수치가 높을수록 다른 작업에 사용할 수 있는 주의가 줄어듭니다.
설치
원클릭 설치
`claude_desktop_config.json` 파일에 다음을 추가하세요:
{
"mcpServers": {
"roundtable": {
"url": "https://mcp.roundtable.now/mcp"
}
}
}원격 엔드포인트
https://mcp.roundtable.now/mcpstreamable-http할 수 있는 일
도구 목록
도구 (13)
🟢 읽기 전용🟡 쓰기🔴 삭제⚪ 알 수 없음
🟢list-models(thinking_level)
List available AI models grouped by thinking level (low/medium/high). Shows default models, credit costs, capabilities for each tier. Use this before consult to understand model options.
입력 스키마
{
"type": "object",
"properties": {
"thinking_level": {
"description": "Filter to a specific thinking level",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"levels": {
"type": "array",
"items": {
"type": "object",
"properties": {
"label": {
"type": "string",
"description": "Human-readable tier name"
},
"level": {
"type": "string",
"description": "Thinking level identifier"
},
"models": {
"type": "array",
"items": {
"type": "object",
"properties": {
"capabilities": {
"type": "array",
"items": {
"type": "string"
},
"description": "Model capabilities"
},
"id": {
"type": "string",
"description": "Model identifier"
},
"multiplier": {
"type": "number",
"description": "Credit cost multiplier"
}
},
"required": [
"capabilities",
"id",
"multiplier"
],
"additionalProperties": false,
"description": "Available model"
},
"description": "Models in this tier"
},
"moderator": {
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Moderator model identifier"
},
"multiplier": {
"type": "number",
"description": "Credit cost multiplier"
}
},
"required": [
"id",
"multiplier"
],
"additionalProperties": false,
"description": "Moderator model for this tier"
}
},
"required": [
"label",
"level",
"models",
"moderator"
],
"additionalProperties": false,
"description": "Thinking level tier"
},
"description": "Available thinking level tiers"
}
},
"required": [
"levels"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟢list-sessions(limit, offset, tool_name)
List your previous MCP tool sessions. Returns session metadata including prompt, tool used, quality score, and credits consumed. Useful for reviewing past council discussions.
입력 스키마
{
"type": "object",
"properties": {
"limit": {
"default": 20,
"description": "Max results to return",
"type": "integer",
"minimum": 1,
"maximum": 100
},
"offset": {
"default": 0,
"description": "Pagination offset",
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
},
"tool_name": {
"description": "Filter by tool name (e.g., \"consult\", \"architect\")",
"type": "string"
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"hasMore": {
"type": "boolean",
"description": "Whether more results are available"
},
"sessions": {
"type": "array",
"items": {
"type": "object",
"properties": {
"credits": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "Credits consumed"
},
"durationMs": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "Duration in milliseconds"
},
"id": {
"type": "string",
"description": "Session identifier"
},
"prompt": {
"type": "string",
"description": "Original prompt text"
},
"thinkingLevel": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Thinking level used"
},
"toolName": {
"type": "string",
"description": "Tool that was invoked"
}
},
"required": [
"credits",
"durationMs",
"id",
"prompt",
"thinkingLevel",
"toolName"
],
"additionalProperties": false,
"description": "Session entry"
},
"description": "List of sessions"
}
},
"required": [
"hasMore",
"sessions"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟢get-session(session_id)
Get full details of a previous MCP session by ID. Returns the complete result including participant responses and moderator synthesis. Use list-sessions first to find session IDs.
입력 스키마
{
"type": "object",
"properties": {
"session_id": {
"type": "string",
"minLength": 1,
"description": "The session ID to retrieve"
}
},
"required": [
"session_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"createdAt": {
"type": "string",
"description": "Session creation timestamp"
},
"durationMs": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "Duration in milliseconds"
},
"id": {
"type": "string",
"description": "Session identifier"
},
"prompt": {
"type": "string",
"description": "Original prompt text"
},
"qualityScore": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "AI-rated quality score"
},
"toolName": {
"type": "string",
"description": "Tool that was invoked"
},
"totalCredits": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"description": "Total credits consumed"
}
},
"required": [
"createdAt",
"durationMs",
"id",
"prompt",
"qualityScore",
"toolName",
"totalCredits"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟢get-logs(end_time, event, level, limit, offset, ...)
Query structured logs from your MCP tool executions. Filter by session, severity level, event type, and time range. Useful for debugging and monitoring tool usage.
입력 스키마
{
"type": "object",
"properties": {
"end_time": {
"description": "End timestamp (ms) for time range filter",
"type": "number"
},
"event": {
"description": "Filter by event name (e.g., \"debate_completed\")",
"type": "string"
},
"level": {
"description": "Filter by log level",
"type": "string",
"enum": [
"info",
"warn",
"error"
]
},
"limit": {
"default": 50,
"description": "Max results",
"type": "integer",
"minimum": 1,
"maximum": 200
},
"offset": {
"default": 0,
"description": "Pagination offset",
"type": "integer",
"minimum": 0,
"maximum": 9007199254740991
},
"session_id": {
"description": "Filter logs for a specific session",
"type": "string"
},
"start_time": {
"description": "Start timestamp (ms) for time range filter",
"type": "number"
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"logs": {
"type": "array",
"items": {
"type": "object",
"properties": {
"data": {
"anyOf": [
{
"type": "object",
"propertyNames": {
"type": "string"
},
"additionalProperties": {
"anyOf": [
{
"type": "string"
},
{
"type": "number"
},
{
"type": "boolean"
},
{
"type": "null"
}
]
}
},
{
"type": "null"
}
],
"description": "Structured log data payload"
},
"event": {
"type": "string",
"description": "Event name"
},
"level": {
"type": "string",
"description": "Log severity level"
},
"sessionId": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Associated session identifier"
},
"time": {
"type": "string",
"description": "Log timestamp"
}
},
"required": [
"data",
"event",
"level",
"sessionId",
"time"
],
"additionalProperties": false,
"description": "Log entry"
},
"description": "Matching log entries"
}
},
"required": [
"logs"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟢check-usage(verbose)
Check your remaining credits, usage limits, and plan info
입력 스키마
{
"type": "object",
"properties": {
"verbose": {
"description": "Include detailed per-window rate limit breakdown",
"default": false,
"type": "boolean"
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"credits": {
"type": "number",
"description": "Remaining credit balance"
},
"plan": {
"type": "string",
"description": "Current plan tier"
},
"rateLimits": {
"type": "object",
"properties": {
"daily": {
"type": "object",
"properties": {
"count": {
"type": "number",
"description": "Requests used"
},
"limit": {
"type": "number",
"description": "Request cap"
},
"resetAt": {
"type": "string",
"description": "Reset timestamp"
}
},
"required": [
"count",
"limit",
"resetAt"
],
"additionalProperties": false,
"description": "Daily rate limit window"
},
"fiveHour": {
"type": "object",
"properties": {
"count": {
"type": "number",
"description": "Requests used"
},
"limit": {
"type": "number",
"description": "Request cap"
},
"resetAt": {
"type": "string",
"description": "Reset timestamp"
}
},
"required": [
"count",
"limit",
"resetAt"
],
"additionalProperties": false,
"description": "5-hour rate limit window"
},
"weekly": {
"type": "object",
"properties": {
"count": {
"type": "number",
"description": "Requests used"
},
"limit": {
"type": "number",
"description": "Request cap"
},
"resetAt": {
"type": "string",
"description": "Reset timestamp"
}
},
"required": [
"count",
"limit",
"resetAt"
],
"additionalProperties": false,
"description": "Weekly rate limit window"
}
},
"required": [
"daily",
"fiveHour",
"weekly"
],
"additionalProperties": false,
"description": "Rate limit status across all windows"
},
"status": {
"type": "string",
"description": "Overall usage status"
}
},
"required": [
"credits",
"plan",
"rateLimits",
"status"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟢get-thread-link(session_id)
Get the dashboard URL for a previous debate session. Returns the thread link and public URL if the thread is public.
입력 스키마
{
"type": "object",
"properties": {
"session_id": {
"type": "string",
"minLength": 1,
"description": "The session ID to get the thread link for"
}
},
"required": [
"session_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"dashboardUrl": {
"type": "string",
"description": "Dashboard URL to view the thread"
},
"isPublic": {
"type": "boolean",
"description": "Whether the thread is publicly accessible"
},
"publicUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Public share URL if thread is public"
}
},
"required": [
"dashboardUrl",
"isPublic",
"publicUrl"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}🟡set-thread-visibility(is_public, session_id)
Set a thread as public or private. Public threads can be shared via URL. Use session_id from a previous debate session.
입력 스키마
{
"type": "object",
"properties": {
"is_public": {
"type": "boolean",
"description": "Set to true to make the thread publicly accessible, false to make it private"
},
"session_id": {
"type": "string",
"minLength": 1,
"description": "The session ID from a previous debate"
}
},
"required": [
"is_public",
"session_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}출력 스키마
{
"type": "object",
"properties": {
"isPublic": {
"type": "boolean",
"description": "Current visibility state"
},
"publicUrl": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Public share URL if thread is public"
}
},
"required": [
"isPublic",
"publicUrl"
],
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false
}⚪consult-council(context, format, mode, models, prompt, ...)
Consult the AI coding council — multiple models discuss your engineering question sequentially (each sees prior responses), then a moderator synthesizes. Auto-mode by default — AI picks optimal models, roles, and conversation mode from your prompt. Provide explicit models to override (manual mode). Fully configurable: mode, format, roles, models, thinking level.
입력 스키마
{
"type": "object",
"properties": {
"context": {
"description": "Additional background context for the debate (code, docs, requirements)",
"type": "string",
"maxLength": 50000
},
"format": {
"default": "discussion",
"description": "Moderator output format: discussion (narrative), adr (architecture decision), comparison (table), pros-cons",
"type": "string",
"enum": [
"discussion",
"adr",
"comparison",
"pros-cons"
]
},
"mode": {
"default": "debating",
"description": "Conversation mode: analyzing (research), brainstorming (ideas), debating (tradeoffs), solving (action plans)",
"type": "string",
"enum": [
"analyzing",
"brainstorming",
"debating",
"solving"
]
},
"models": {
"description": "Override specific model IDs. Min 3 models. Use list-models to see available options",
"minItems": 3,
"maxItems": 6,
"type": "array",
"items": {
"type": "string"
}
},
"prompt": {
"type": "string",
"minLength": 1,
"maxLength": 10000,
"description": "The question, topic, or problem to debate"
},
"roles": {
"description": "Inline role names for participants (e.g., [\"Security Architect\", \"Backend Engineer\"])",
"type": "array",
"items": {
"type": "string"
}
},
"thinking_level": {
"default": "medium",
"description": "Controls model quality and cost: low (fast/cheap), medium (balanced), high (maximum reasoning)",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
},
"session_context": {
"description": "Session IDs to use as context (max 3). Prior moderator summaries will be prepended.",
"maxItems": 3,
"type": "array",
"items": {
"type": "string"
}
},
"knowledge": {
"description": "Reference knowledge to inject as context",
"maxItems": 5,
"type": "array",
"items": {
"type": "object",
"properties": {
"content": {
"type": "string",
"description": "Text content or URL to fetch"
},
"label": {
"description": "Label for this knowledge item",
"type": "string"
},
"type": {
"type": "string",
"enum": [
"text",
"url"
],
"description": "Content type: text (pass-through) or url (fetch and extract)"
}
},
"required": [
"content",
"type"
]
}
},
"auto_route": {
"default": false,
"description": "Auto-select optimal models based on prompt analysis and historical performance",
"type": "boolean"
}
},
"required": [
"prompt"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪design-architecture(description, focus_areas, scale, tech_stack, webhook_url)
Architecture design council. Systems Architect, Infrastructure Engineer, and DX Advocate evaluate your system design. Always uses high thinking for maximum depth. Output as ADR.
입력 스키마
{
"type": "object",
"properties": {
"description": {
"type": "string",
"minLength": 1,
"maxLength": 10000,
"description": "What the system should do"
},
"focus_areas": {
"description": "Priority areas (e.g., [\"security\", \"performance\"])",
"type": "array",
"items": {
"type": "string"
}
},
"scale": {
"default": "startup",
"description": "Target scale: startup (small team), growth (scaling), enterprise (large org)",
"type": "string",
"enum": [
"startup",
"growth",
"enterprise"
]
},
"tech_stack": {
"description": "Preferred technologies",
"type": "array",
"items": {
"type": "string"
}
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
}
},
"required": [
"description"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢review-code(code, focus, language, thinking_level, webhook_url)
Code review council. Senior Engineer, Security Reviewer, and Performance Analyst analyze your code and a moderator synthesizes their findings.
입력 스키마
{
"type": "object",
"properties": {
"code": {
"type": "string",
"minLength": 1,
"maxLength": 100000,
"description": "The code to review"
},
"focus": {
"description": "Review focus areas (e.g., [\"security\", \"performance\"])",
"type": "array",
"items": {
"type": "string"
}
},
"language": {
"description": "Programming language (auto-detected if not specified)",
"type": "string"
},
"thinking_level": {
"default": "medium",
"description": "Review depth: low (quick scan), medium (balanced), high (thorough)",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
}
},
"required": [
"code"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪plan-implementation(codebase_context, constraints, feature, tech_stack, thinking_level, ...)
Implementation planning council. Tech Lead, Senior Engineer, and QA Strategist break down a feature into actionable steps, identify risks, and define acceptance criteria. Output as ADR.
입력 스키마
{
"type": "object",
"properties": {
"codebase_context": {
"description": "Relevant existing code, file structure, or architecture notes",
"type": "string",
"maxLength": 50000
},
"constraints": {
"description": "Constraints (e.g., [\"no breaking changes\", \"must support offline\"])",
"type": "array",
"items": {
"type": "string"
}
},
"feature": {
"type": "string",
"minLength": 1,
"maxLength": 10000,
"description": "The feature or change to plan"
},
"tech_stack": {
"description": "Current tech stack",
"type": "array",
"items": {
"type": "string"
}
},
"thinking_level": {
"default": "medium",
"description": "Planning depth",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
},
"session_context": {
"description": "Session IDs to use as context (max 3). Prior moderator summaries will be prepended.",
"maxItems": 3,
"type": "array",
"items": {
"type": "string"
}
},
"knowledge": {
"description": "Reference knowledge to inject as context",
"maxItems": 5,
"type": "array",
"items": {
"type": "object",
"properties": {
"content": {
"type": "string",
"description": "Text content or URL to fetch"
},
"label": {
"description": "Label for this knowledge item",
"type": "string"
},
"type": {
"type": "string",
"enum": [
"text",
"url"
],
"description": "Content type: text (pass-through) or url (fetch and extract)"
}
},
"required": [
"content",
"type"
]
}
}
},
"required": [
"feature"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢debug-issue(code, error, expected_behavior, problem, thinking_level, ...)
Debugging council. Root Cause Analyst, Systems Engineer, and Edge Case Investigator collaboratively diagnose bugs, analyze errors, and propose fixes.
입력 스키마
{
"type": "object",
"properties": {
"code": {
"description": "The relevant code where the bug occurs",
"type": "string",
"maxLength": 100000
},
"error": {
"description": "Error message, stack trace, or unexpected output",
"type": "string",
"maxLength": 10000
},
"expected_behavior": {
"description": "What should happen vs what actually happens",
"type": "string",
"maxLength": 5000
},
"problem": {
"type": "string",
"minLength": 1,
"maxLength": 10000,
"description": "Describe the bug, failure, or unexpected behavior"
},
"thinking_level": {
"default": "medium",
"description": "Analysis depth",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
},
"session_context": {
"description": "Session IDs to use as context (max 3). Prior moderator summaries will be prepended.",
"maxItems": 3,
"type": "array",
"items": {
"type": "string"
}
},
"knowledge": {
"description": "Reference knowledge to inject as context",
"maxItems": 5,
"type": "array",
"items": {
"type": "object",
"properties": {
"content": {
"type": "string",
"description": "Text content or URL to fetch"
},
"label": {
"description": "Label for this knowledge item",
"type": "string"
},
"type": {
"type": "string",
"enum": [
"text",
"url"
],
"description": "Content type: text (pass-through) or url (fetch and extract)"
}
},
"required": [
"content",
"type"
]
}
}
},
"required": [
"problem"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪assess-tradeoffs(context, decision, options, priorities, thinking_level, ...)
Tradeoff assessment council. Pragmatist, Skeptic, and Futurist evaluate options from different angles — short-term vs long-term, risk vs reward, simplicity vs flexibility. Output as pros-cons.
입력 스키마
{
"type": "object",
"properties": {
"context": {
"description": "Background context — codebase, team, timeline, constraints",
"type": "string",
"maxLength": 50000
},
"decision": {
"type": "string",
"minLength": 1,
"maxLength": 10000,
"description": "The decision or question to evaluate"
},
"options": {
"description": "Specific options to compare",
"minItems": 2,
"type": "array",
"items": {
"type": "string"
}
},
"priorities": {
"description": "What matters most (e.g., [\"performance\", \"dx\", \"cost\"])",
"type": "array",
"items": {
"type": "string"
}
},
"thinking_level": {
"default": "medium",
"description": "Analysis depth",
"type": "string",
"enum": [
"low",
"medium",
"high"
]
},
"webhook_url": {
"description": "Webhook URL to POST results to after completion",
"type": "string",
"format": "uri"
},
"session_context": {
"description": "Session IDs to use as context (max 3). Prior moderator summaries will be prepended.",
"maxItems": 3,
"type": "array",
"items": {
"type": "string"
}
}
},
"required": [
"decision"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}권장 프롬프트
retrieve_data
Get details about [item] from Roundtable
예상 도구:
get-sessionfetch_info
Fetch [information type] using Roundtable
예상 도구:
get-sessionlist_items
List all [items] available in Roundtable
예상 도구:
list-modelsbrowse_collection
Show me the [collection] from Roundtable
예상 도구:
list-modelsexplore_workflow
List available [items], then get details for each one using Roundtable
예상 도구:
list-modelsget-session커뮤니티
증거
최근 관측
검증됨버전이 기록되지 않음도구 13개
검증됨버전이 기록되지 않음도구 13개