Skip to main content
Glama
ZephyrDeng

mcp-server-gitlab

Gitlab Create MR Tool

Generate new Merge Requests in GitLab, assign users, and add reviewers for streamlined code collaboration. Simplify project management by specifying branches, titles, and labels.

Instructions

创建新的 Merge Request,支持指派 assignee 和 reviewers。

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
assigneeIdNo指派的用户 ID
descriptionNo描述
fieldsNo需要返回的字段路径数组
labelsNo标签数组
projectIdYes项目 ID
reviewerIdsNoReviewer 用户 ID 列表
sourceBranchYes源分支
targetBranchYes目标分支
titleYes标题

Implementation Reference

  • Main tool handler: the `execute` method that performs project/user resolution, GitLab API POST to create merge request, response filtering, and error handling.
    async execute(args: unknown, context: Context<Record<string, unknown> | undefined>) {
      const typedArgs = args as {
        projectId: string | number;
        sourceBranch: string;
        targetBranch: string;
        title: string;
        description?: string;
        assigneeId?: string | number;
        reviewerIds?: (string | number)[];
        labels?: string[];
        fields?: string[];
      };
    
      const { projectId: projectIdOrName, sourceBranch, targetBranch, title, description, assigneeId: assigneeIdOrName, reviewerIds: reviewerIdsOrNames, labels, fields } = typedArgs;
    
      try {
        const client = createGitlabClientFromContext(context);
        // 解析项目 ID
        const resolvedProjectId = await client.resolveProjectId(projectIdOrName);
        if (!resolvedProjectId) {
          throw new Error(`无法解析项目 ID 或名称:${projectIdOrName}`);
        }
        const endpoint = `/projects/${encodeURIComponent(String(resolvedProjectId))}/merge_requests`;
    
        let resolvedAssigneeId: number | undefined = undefined;
        if (assigneeIdOrName !== undefined) {
          const resolved = await client.resolveUserId(assigneeIdOrName);
          if (resolved) {
            resolvedAssigneeId = resolved;
          } else {
            // 选择性失败:可以记录警告并继续,或者抛出错误
             console.warn(`无法解析指派人 ID 或用户名:${assigneeIdOrName},将忽略此参数`);
            // throw new Error(`无法解析指派人 ID 或用户名:${assigneeIdOrName}`);
          }
        }
    
        let resolvedReviewerIds: number[] = [];
        if (reviewerIdsOrNames && reviewerIdsOrNames.length > 0) {
          const reviewerPromises = reviewerIdsOrNames.map(idOrName => client.resolveUserId(idOrName));
          const results = await Promise.all(reviewerPromises);
          resolvedReviewerIds = results.filter((id): id is number => {
            if (id === null) {
              // 找到原始输入用于警告
              const originalInput = reviewerIdsOrNames[results.indexOf(id)];
              console.warn(`无法解析审阅人 ID 或用户名:${originalInput},将忽略此审阅人`);
            }
            return id !== null;
          });
        }
    
    
        const createData: Record<string, any> = {
          source_branch: sourceBranch,
          target_branch: targetBranch,
          title,
        };
        if (description !== undefined) createData.description = description;
        if (resolvedAssigneeId !== undefined) createData.assignee_id = resolvedAssigneeId;
        if (resolvedReviewerIds.length > 0) createData.reviewer_ids = resolvedReviewerIds;
        if (labels !== undefined) createData.labels = labels.join(",");
    
    
        const response = await client.apiRequest(endpoint, "POST", undefined, createData);
    
        if (!client.isValidResponse(response)) {
          throw new Error(`GitLab API error: ${response?.message || 'Unknown error'}`);
        }
    
        if (fields) {
          const filteredResponse = filterResponseFields(response, fields);
          return {
            content: [{ type: "text", text: JSON.stringify(filteredResponse) }]
          } as ContentResult;
        }
    
        return {
          content: [{ type: "text", text: JSON.stringify(response) }]
        } as ContentResult;
      } catch (error: any) {
        return {
          content: [
            {
              type: "text",
              text: `GitLab MCP 工具调用异常:${error?.message || String(error)}`
            }
          ],
          isError: true
        };
      }
    }
  • Zod schema defining the tool's input parameters including projectId, branches, title, assignees, reviewers, etc.
    parameters: z.object({
      projectId: z.union([z.string(), z.number()]).describe("项目 ID 或名称"),
      sourceBranch: z.string().describe("源分支"),
      targetBranch: z.string().describe("目标分支"),
      title: z.string().describe("标题"),
      description: z.string().optional().describe("描述"),
      assigneeId: z.union([z.string(), z.number()]).optional().describe("指派的用户 ID 或用户名"),
      reviewerIds: z.array(z.union([z.string(), z.number()])).optional().describe("Reviewer 用户 ID 或用户名列表"),
      labels: z.array(z.string()).optional().describe("标签数组"),
      fields: z.array(z.string()).optional().describe("需要返回的字段路径数组"),
    }),
  • Inclusion of `GitlabCreateMRTool` in the `fastmcpTools` array, which is iterated over to register all GitLab tools with the MCP server.
    const fastmcpTools = [
      GitlabAcceptMRTool,
      GitlabCreateMRCommentTool,
      GitlabCreateMRTool,
      GitlabGetUserTasksTool,
      GitlabRawApiTool,
      GitlabSearchProjectDetailsTool,
      GitlabSearchUserProjectsTool,
      GitlabUpdateMRTool,
    ];
  • The registration loop in `registerGitLabToolsForFastMCP` that conditionally adds each tool to the FastMCP server using `server.addTool(tool)`. A similar process exists for MCP SDK registration.
    fastmcpTools.forEach(tool => {
      const standardizedName = toolNameMapping[tool.name as keyof typeof toolNameMapping];
      if (shouldRegisterTool(standardizedName as GitLabToolName, options.toolFilter)) {
        // GitLabTool is now fully compatible with FastMCP's base type, can be registered directly
        server.addTool(tool);
      }
    });
  • Mapping of the tool's display name to the standardized `GitLabToolName` enum value 'Gitlab_Create_MR_Tool' for type safety and filtering.
    [GitlabCreateMRTool.name]: "Gitlab_Create_MR_Tool",
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. While '创建新的 Merge Request' implies a write/mutation operation, the description lacks critical details: it doesn't mention authentication requirements, potential side effects (e.g., branch merges), error conditions, or what the tool returns. For a creation tool with zero annotation coverage, this leaves significant gaps in understanding its behavior.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is extremely concise—a single sentence in Chinese that directly states the tool's core function and key features. It's front-loaded with the primary action and wastes no words, making it easy to parse quickly. Every part of the sentence earns its place by highlighting essential capabilities.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity of creating a Merge Request (a mutation with 9 parameters), no annotations, and no output schema, the description is incomplete. It doesn't explain the return value, error handling, or important behavioral aspects like required permissions or idempotency. While the schema covers parameters well, the overall context for safe and effective use is lacking.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all 9 parameters thoroughly. The description adds minimal value beyond the schema by mentioning assignee and reviewer support, which aligns with the 'assigneeId' and 'reviewerIds' parameters but doesn't provide additional context like format examples or constraints. This meets the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: '创建新的 Merge Request' (create a new Merge Request) with specific capabilities: '支持指派 assignee 和 reviewers' (supports assigning assignee and reviewers). This is a specific verb+resource combination that distinguishes it from siblings like Gitlab Update MR Tool or Gitlab Accept MR Tool. However, it doesn't explicitly differentiate from Gitlab Create MR Comment Tool beyond the core action.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives. It doesn't mention prerequisites (e.g., needing existing branches), compare it to Gitlab Update MR Tool for modifications, or specify scenarios where it's appropriate. The agent must infer usage from the name and parameters alone, which is insufficient for clear decision-making.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Related Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ZephyrDeng/mcp-server-gitlab'

If you have feedback or need assistance with the MCP directory API, please join our Discord server