Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault

No arguments

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Server capabilities have not been inspected yet.

Tools

Functions exposed to the LLM to take actions

NameDescription
mmla_validate_codeA

Validate code against MMLA specification.

🚨 核心修正 1: 絕對門禁檢查 只有狀態為 GREEN 的節點才能進行代碼驗證

Args: code: The Python code to validate node_id: The MMLA node ID to validate against use_agentic_loop: If True, use Agentic Loop with auto-fix (up to 16 retries)

Returns: JSON string with validation results

mmla_update_statusD
mmla_create_nodeB

Create a new node in the architecture. 'spec' should be a JSON string defining inputs/outputs/constraints.

check_bluemouse_environmentA

檢查藍圖小老鼠運行環境

檢測宿主環境(Antigravity/Cursor/VSCode)、API Key配置和依賴狀態。

Returns: JSON 格式的環境檢測報告

open_bluemouse_uiC

啟動藍圖小老鼠 UI

在瀏覽器中打開藍圖小老鼠的用戶界面,開始使用者旅程。

Args: api_key: 可選的 API Key(BYOK模式) mode: 啟動模式 ("landing" | "workspace")

Returns: UI 啟動狀態和 URL

analyze_requirement_trapA

分析用戶需求並檢測是否需要蘇格拉底面試

檢測需求中的模糊點、邏輯漏洞和潛在災難場景, 如果發現問題則自動生成蘇格拉底式問題。

Args: user_input: 用戶的系統需求描述

Returns: JSON 格式的分析結果,包含是否需要面試和問題列表

record_socratic_answersC

記錄蘇格拉底面試的答案到 data_trap.jsonl

用於訓練數據收集(如果用戶允許)。

Args: requirement: 原始需求 questions: 問題列表(JSON字符串) answers: 用戶答案(JSON字符串) framework: 選擇的框架

Returns: 記錄狀態

deliver_bluemouse_projectC

將生成的項目文件寫入宿主工作區

完成從「寄生」到「交付」的完整閉環。

Args: project_name: 項目名稱 files: 文件映射 (JSON字符串) metadata: 元數據 (JSON字符串)

Returns: 生成報告

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription
get_summary

TDQS

C2.4/5.0

Scored across 8 tools

Disambiguation3/5

The tool set has clear distinct purposes for most tools, but there is some overlap between analyze_requirement_trap and record_socratic_answers in handling Socratic interview data, which could cause confusion. The MMLA tools (mmla_create_node, mmla_update_status, mmla_validate_code) are well-differentiated from the BlueMouse environment tools, but the lack of a description for mmla_update_status adds ambiguity. Overall, descriptions help clarify, but some boundaries are fuzzy.

Naming Consistency2/5

Naming is inconsistent with mixed conventions: snake_case (e.g., analyze_requirement_trap, check_bluemouse_environment) is used for most tools, but mmla_create_node uses a prefix with snake_case, and mmla_update_status and mmla_validate_code follow a similar pattern but lack uniformity in verb usage. There is no clear overall pattern, making it harder to predict tool names or their purposes based on naming alone.

Tool Count4/5

With 8 tools, the count is reasonable for a server focused on project analysis, environment management, and MMLA architecture. It covers core workflows without being overly heavy, though it might feel slightly thin if expanded to more complex domains. The number aligns well with the apparent scope of BlueMouse's functionality.

Completeness3/5

The tool surface covers key areas like requirement analysis, environment checks, project delivery, MMLA node management, and UI interaction, but there are notable gaps. For example, there are no tools for updating or deleting MMLA nodes, and the Socratic interview process lacks tools for modifying or reviewing recorded data. This could lead to workarounds or incomplete agent workflows in some scenarios.

Maintenance

ActivityInactive
ResponsivenessNo issues