feat: 支持初步的CR

2024-08-12 12:24:45 +00:00 · 2024-08-12 12:24:45 +00:00 · 92fa30ef3d
commit 92fa30ef3d
parent a72a11fedb
37 changed files with 3925 additions and 15 deletions
--- a/.vscode/settings.json
+++ b/.vscode/settings.json
@ -8,12 +8,17 @@
    "cloudml",
    "commitlint",
    "dbaeumer",
+    "deepseek",
    "devcontainers",
    "eamodio",
    "esbenp",
+    "gitbeaker",
    "Gruntfuggly",
+    "Hasher",
+    "langchain",
    "micr",
    "mioffice",
+    "openai",
    "oxlint",
    "tseslint",
    "wlpbbgiky",
--- a/bun.lockb
+++ b/bun.lockb
--- a/controllers/manageMrEvent/index.ts
+++ b/controllers/manageMrEvent/index.ts
@ -0,0 +1,152 @@
+import { Logger } from "winston"
+
+import service from "../../service"
+import netTool from "../../service/netTool"
+import { Gitlab } from "../../types/gitlab"
+import reviewFiles from "./reviewFiles"
+import summaryFiles from "./summaryFiles"
+import summaryMr from "./summaryMr"
+import Commenter from "./utils/commenter"
+import { Inputs } from "./utils/inputs"
+import { PathFilter } from "./utils/pathFilter"
+
+/**
+ * 过滤事件
+ * @param {Gitlab.MergeRequestEvent} event - 合并请求事件
+ * @returns {boolean} - 是否通过过滤
+ */
+const filterEvent = (event: Gitlab.MergeRequestEvent): boolean => {
+  return (
+    event.event_type === "merge_request" &&
+    event.object_attributes?.state === "opened" &&
+    !event.object_attributes?.changes
+  )
+}
+
+/**
+ * 处理原始事件
+ * @param {Gitlab.MergeRequestEvent} mergeRequest - 合并请求事件
+ * @param {Logger} logger - 日志记录器
+ * @returns {Promise<any>} - 处理结果
+ */
+const manageRawEvent = async (
+  mergeRequest: Gitlab.MergeRequestEvent,
+  logger: Logger
+): Promise<any> => {
+  // 如果MR不是打开或重新打开状态，则不处理
+  const isOnCreateMr = filterEvent(mergeRequest)
+  logger.info(`isOnCreateMr: ${isOnCreateMr}`)
+  if (!isOnCreateMr) return netTool.ok()
+  // 获取最新的MR Version
+  const versions = await service.gitlab.mr.getDiffVersions(
+    mergeRequest.project.id,
+    mergeRequest.object_attributes.iid
+  )
+  if (versions.length === 0) {
+    logger.error("Failed to get MR versions")
+    return netTool.serverError("Failed to get MR versions")
+  }
+  const latestVersion = versions[0]
+  logger.debug(`latestVersion: ${JSON.stringify(latestVersion)}`)
+  // 获取MR的全部修改
+  const changes = await service.gitlab.mr.getChanges(
+    mergeRequest.project.id,
+    mergeRequest.object_attributes.iid
+  )
+  logger.debug(`changes: ${JSON.stringify(changes)}`)
+  // 如果MR的全部修改不存在，则不处理
+  if (!changes) {
+    logger.error("Failed to get MR changes")
+    return netTool.serverError("Failed to get MR changes")
+  }
+  // 如果MR的描述中包含'skip review'，则不处理
+  if (changes.description.includes("skip review")) {
+    logger.info("Skip review")
+    return netTool.ok()
+  }
+  // 创建路径过滤器
+  const pathFilter = new PathFilter([
+    "**/*.css",
+    "**/*.less",
+    "**/*.scss",
+    "**/*.js",
+    "**/*.jsx",
+    "**/*.ts",
+    "**/*.tsx",
+  ])
+  // 获取MR的文件修改，并过滤出需要处理的文件
+  const files = changes.changes.filter(
+    (change) =>
+      pathFilter.check(change.new_path) &&
+      !change.deleted_file &&
+      !change.renamed_file &&
+      change.diff
+  )
+  // 如果没有需要处理的文件，则不处理
+  if (files.length === 0) {
+    logger.info("No files need to be processed")
+    return netTool.ok()
+  }
+  // 创建Commenter实例
+  const commenter = new Commenter(
+    latestVersion.base_commit_sha,
+    latestVersion.head_commit_sha,
+    latestVersion.start_commit_sha,
+    mergeRequest.project.id,
+    mergeRequest.object_attributes.iid,
+    logger
+  )
+  // 发起loading评论，无所谓是否成功
+  await commenter.createLoadingComment()
+
+  // 创建输入实例
+  const inputs: Inputs = new Inputs()
+  // 获取MR的标题和描述
+  inputs.title = changes.title
+  inputs.description = changes.description || "暂无描述"
+
+  // 对文件进行总结
+  const { summarizedFileMap, needReviewFileMap } = await summaryFiles(
+    files,
+    true,
+    inputs,
+    logger
+  )
+  // 如果总结完的文件Map为空，则不处理
+  if (summarizedFileMap.size === 0) {
+    logger.info("No summarized files")
+    await commenter.modifyLoadingOrCreateComment("没有需要审查的文件")
+    return netTool.ok()
+  }
+  // 总结MR
+  const summarizedMr = await summaryMr(summarizedFileMap, inputs, logger)
+  // 如果总结MR为空，则不处理
+  if (!summarizedMr) {
+    logger.info("No summarized Mr")
+    await commenter.modifyLoadingOrCreateComment("没有生成整体的总结")
+    return netTool.ok()
+  }
+  // 更新输入实例短总结
+  inputs.shortSummary = summarizedMr
+  // 发送总结评论 + 文件变更
+  await commenter.createSummarizedComment(summarizedMr, summarizedFileMap)
+  // 过滤出需要审查的文件
+  const needReviewFiles = files.filter(
+    (file) => needReviewFileMap.get(file.new_path) || false
+  )
+  // 如果需要审查的文件为空，则不处理
+  if (needReviewFiles.length === 0) {
+    logger.info("No need review files")
+    await commenter.createComment("没有需要审查的文件")
+    return netTool.ok()
+  }
+  // 对需要审查的文件进行审查
+  await reviewFiles(needReviewFiles, inputs, commenter, logger)
+  return netTool.ok()
+}
+
+const manageMrEvent = {
+  manageRawEvent,
+}
+
+export default manageMrEvent
--- a/controllers/manageMrEvent/reviewFiles.ts
+++ b/controllers/manageMrEvent/reviewFiles.ts
@ -0,0 +1,152 @@
+import { CommitDiffSchema } from "@gitbeaker/rest"
+import pLimit from "p-limit"
+import { Logger } from "winston"
+
+import chatTools from "../../utils/chatTools"
+import tokenTools from "../../utils/tokenTools"
+import Commenter from "./utils/commenter"
+import diffTools, { Review } from "./utils/diffTools"
+import { Inputs } from "./utils/inputs"
+import { Prompts } from "./utils/prompts"
+
+/**
+ * 审查需要审查的文件列表，并生成评论。
+ * @param {CommitDiffSchema[]} needReviewFiles - 需要审查的文件列表。
+ * @param {Inputs} rawInput - 原始输入。
+ * @param {Commenter} commenter - 评论工具实例。
+ * @param {Logger} logger - 日志记录器。
+ * @returns {Promise<void>} 返回一个Promise，表示审查过程的完成。
+ */
+const reviewFiles = async (
+  needReviewFiles: CommitDiffSchema[],
+  rawInput: Inputs,
+  commenter: Commenter,
+  logger: Logger
+) => {
+  const prompts = new Prompts()
+
+  // 创建一个Map来存储文件的差异信息
+  const diffsFileMap = new Map<string, [number, number, string][]>()
+
+  // 解析每个文件的差异并存储在diffsFileMap中
+  needReviewFiles.forEach((file) => {
+    const diffs = diffTools.parseFileDiffs(file)
+    const fileDiff = diffsFileMap.get(file.new_path) || []
+    diffsFileMap.set(file.new_path, fileDiff.concat(diffs))
+  })
+
+  logger.debug(`diffsFileMap: ${JSON.stringify([...diffsFileMap])}`)
+
+  // 如果没有需要审查的文件，记录日志并创建评论
+  if (diffsFileMap.size === 0) {
+    logger.info("No files need review")
+    commenter.createComment("没有需要审查的文件")
+    return
+  }
+
+  const skippedFiles: string[] = []
+  const fileCRResultsMap = new Map<string, Review[]>()
+
+  /**
+   * 审查单个文件的差异。
+   * @param {Array} diff - 文件的差异数组。
+   * @param {string} filename - 文件名。
+   * @returns {Promise<void>} 返回一个Promise，表示审查过程的完成。
+   */
+  const doFileReview = async (
+    diff: [number, number, string][],
+    filename: string
+  ) => {
+    logger.info(`Reviewing ${filename}`)
+    const inputs = rawInput.clone()
+    inputs.filename = filename
+    let tokens = tokenTools.getTokenCount(prompts.renderReviewFileDiff(inputs))
+
+    // 计算并打包文件的差异，确保令牌数不超过限制
+    const packedFileCount = diff.findIndex((d, idx) => {
+      const hunk = d[2]
+      if (tokens + tokenTools.getTokenCount(hunk) > 100 * 1000) {
+        logger.info(
+          `only packing ${idx} / ${diff.length} diffs, tokens: ${tokens} / ${100 * 1000}`
+        )
+        return true
+      }
+      inputs.patches += `
+${hunk}
+`
+      tokens += tokenTools.getTokenCount(hunk)
+      return false
+    })
+
+    // 如果没有打包任何差异，跳过该文件
+    if (packedFileCount === 0) {
+      logger.info(`skipping ${filename}`)
+      skippedFiles.push(filename)
+      return
+    }
+
+    // 获取GPT-4模型并生成审查提示
+    const codeChatBot = await chatTools.getGpt4oModel(0)
+    const reviewPrompt = prompts.renderReviewFileDiff(inputs)
+    logger.debug(`reviewPrompt for ${filename}: ${reviewPrompt}`)
+    try {
+      const { content: review } = await codeChatBot.invoke(reviewPrompt)
+      if (!review) throw new Error("Empty review")
+      logger.info(`review for ${filename}: ${review}`)
+
+      // 解析审查结果并过滤需要评论的审查
+      const reviews = diffTools.parseReview(review as string, diff)
+      logger.debug(`reviews for ${filename}: ${JSON.stringify(reviews)}`)
+      const needCommentReviews = reviews.filter(
+        (r) => !r.comment.includes("LGTM")
+      )
+      fileCRResultsMap.set(filename, needCommentReviews)
+    } catch {
+      logger.error(`Failed to review ${filename}`)
+    }
+  }
+
+  /**
+   * 创建评论。
+   * @param {string} filename - 文件名。
+   * @param {Review} reviews - 审查结果。
+   * @returns {Promise<void>} 返回一个Promise，表示评论过程的完成。
+   */
+  const doComment = async (filename: string, reviews: Review) => {
+    const file = needReviewFiles.find((f) => f.new_path === filename)
+    if (!file) {
+      logger.error(`Failed to find file ${filename}`)
+      return
+    }
+    await commenter.createReviewComment(
+      file.new_path,
+      file.old_path,
+      reviews.startLine,
+      reviews.endLine,
+      reviews.comment
+    )
+  }
+
+  // 使用pLimit限制并发审查文件的数量
+  const limit = pLimit(10)
+  await Promise.allSettled(
+    [...diffsFileMap].map(([filename, diffs]) =>
+      limit(() => doFileReview(diffs, filename))
+    )
+  )
+
+  // 把fileCRResultsMap中的needCommentReviews拍平成一个数组
+  const needCommentReviews: { filename: string; reviews: Review }[] = []
+  fileCRResultsMap.forEach((reviews, filename) => {
+    reviews.forEach((r) => needCommentReviews.push({ filename, reviews: r }))
+  })
+
+  // 创建评论
+  await Promise.allSettled(
+    needCommentReviews.map(({ filename, reviews }) =>
+      limit(() => doComment(filename, reviews))
+    )
+  )
+}
+
+export default reviewFiles
--- a/controllers/manageMrEvent/summaryFiles.ts
+++ b/controllers/manageMrEvent/summaryFiles.ts
@ -0,0 +1,126 @@
+import { CommitDiffSchema } from "@gitbeaker/rest"
+import pLimit from "p-limit"
+import { Logger } from "winston"
+
+import chatTools from "../../utils/chatTools"
+import tokenTools from "../../utils/tokenTools"
+import { Inputs } from "./utils/inputs"
+import { Prompts } from "./utils/prompts"
+
+/**
+ * 对文件进行总结
+ * @param {CommitDiffSchema[]} files - 要总结的文件
+ * @param {boolean} needReview - 是否审查简单更改
+ * @param {Inputs} rawInputs - 原始输入
+ * @param {Prompts} prompts - 提示
+ * @param {Logger} logger - 日志记录器
+ * @returns {Promise<{ summarizedFileMap: Map<string, string>, needReviewFileMap: Map<string, boolean> }>} 返回包含总结和需要审查的文件Map的Promise
+ */
+const summaryFiles = async (
+  files: CommitDiffSchema[],
+  needReview: boolean,
+  rawInputs: Inputs,
+  logger: Logger
+) => {
+  // 生成Prompts实例
+  const prompts: Prompts = new Prompts()
+  // 按文件名融合Diff
+  const fileMap = new Map<string, string>()
+  const DIFF_SEPARATOR = "\n--- DIFF SEPARATOR ---\n"
+
+  files.forEach((file) => {
+    const rawDiff = fileMap.get(file.new_path) || ""
+    // 融合Diff
+    const diff = rawDiff ? rawDiff + DIFF_SEPARATOR + file.diff : file.diff
+    fileMap.set(file.new_path, diff)
+  })
+
+  // 总结后的文件Map
+  const summarizedFileMap = new Map<string, string>()
+
+  /**
+   * 生成文件的Summary
+   * @param {string} diff - 文件的差异
+   * @param {string} path - 文件路径
+   * @returns {Promise<void>} 返回生成Summary的Promise
+   */
+  const doFileSummary = async (diff: string, path: string) => {
+    const inputs = rawInputs.clone()
+    inputs.fileDiff = diff
+    inputs.filename = path
+    const summarizePrompt = prompts.renderSummarizeFileDiff(inputs, needReview)
+    logger.debug(`summarizePrompt for ${path}: ${summarizePrompt}`)
+    const tokens = tokenTools.getTokenCount(summarizePrompt)
+    logger.debug(`tokens for ${path}: ${tokens}`)
+    if (tokens > 100 * 1000) {
+      logger.error(
+        `File diff too long for ${path} (${tokens} tokens), skipping`
+      )
+      return
+    }
+    const codeChatBot = await chatTools.getGpt4oModel(0)
+    try {
+      const { content: summarize } = await codeChatBot.invoke(summarizePrompt)
+      if (!summarize) throw new Error("Empty summarize")
+      logger.info(`summarize for ${path}: ${summarize}`)
+      summarizedFileMap.set(path, summarize as string)
+    } catch {
+      logger.error(`Failed to summarize for ${path}`)
+    }
+  }
+
+  const limit = pLimit(5)
+  const promises = Array.from(fileMap.entries()).map(([path, diff]) =>
+    limit(() => doFileSummary(diff, path))
+  )
+  await Promise.allSettled(promises)
+
+  // 需要Review的文件Map
+  const needReviewFileMap = new Map<string, boolean>()
+
+  // 如果不需要审查更改，则直接返回
+  if (!needReview)
+    return {
+      summarizedFileMap,
+      needReviewFileMap,
+    }
+
+  /**
+   * 管理文件的审查状态
+   * @param {string} path - 文件路径
+   * @param {string} summarize - 文件总结
+   */
+  const manageTriage = (path: string, summarize: string) => {
+    const triageRegex = /\[TRIAGE\]:\s*(NEEDS_REVIEW|APPROVED)/
+    const triageMatch = summarize.match(triageRegex)
+    // 如果没有匹配到TRIAGE，打印错误日志
+    if (!triageMatch) {
+      logger.error(`Failed to triage for ${path}`)
+      needReviewFileMap.set(path, true)
+      return
+    }
+    // 如果匹配到TRIAGE，根据匹配结果设置needReviewFileMap
+    if (triageMatch[1] === "APPROVED") {
+      logger.info(`Approved for ${path}`)
+      needReviewFileMap.set(path, false)
+    } else {
+      logger.info(`Needs review for ${path}`)
+      needReviewFileMap.set(path, true)
+    }
+    // 删除源总结中的TRIAGE
+    const newSummarize = summarize.replace(triageRegex, "").trim()
+    summarizedFileMap.set(path, newSummarize)
+  }
+
+  // 全部文件过一遍TRIAGE
+  Array.from(summarizedFileMap.entries()).forEach(([path, summarize]) =>
+    manageTriage(path, summarize)
+  )
+
+  return {
+    summarizedFileMap,
+    needReviewFileMap,
+  }
+}
+
+export default summaryFiles
--- a/controllers/manageMrEvent/summaryMr.ts
+++ b/controllers/manageMrEvent/summaryMr.ts
@ -0,0 +1,56 @@
+import { Logger } from "winston"
+
+import chatTools from "../../utils/chatTools"
+import { Inputs } from "./utils/inputs"
+import { Prompts } from "./utils/prompts"
+
+/**
+ * 总结合并请求的变更。
+ * @param {Map<string, string>} summarizedFileMap - 一个包含文件路径和对应摘要的Map。
+ * @param {Inputs} rawInputs - 输入数据的实例。
+ * @param {Logger} logger - 用于记录日志的Logger实例。
+ * @returns {Promise<string>} 返回包含总结的Promise字符串。
+ */
+const summaryMr = async (
+  summarizedFileMap: Map<string, string>,
+  rawInputs: Inputs,
+  logger: Logger
+) => {
+  // 创建Prompts实例
+  const prompts = new Prompts()
+  // 克隆输入数据
+  const inputs = rawInputs.clone()
+
+  // 遍历summarizedFileMap，将每个文件的摘要添加到rawSummary中
+  summarizedFileMap.forEach((summarize, path) => {
+    inputs.rawSummary += `---
+    ${path}: ${summarize}
+    `
+  })
+  // 记录完整的变更日志
+  logger.debug(`full changes: ${inputs.rawSummary}`)
+
+  // 渲染总结提示
+  const summarizePrompt = prompts.renderSummarize(inputs)
+
+  // 记录总结提示日志
+  logger.debug(`summarizePrompt for Mr: ${summarizePrompt}`)
+
+  // 获取GPT-4o模型实例
+  const codeChatBot = await chatTools.getGpt4oModel(0)
+
+  try {
+    // 调用模型生成总结
+    const { content: summarize } = await codeChatBot.invoke(summarizePrompt)
+    if (!summarize) throw new Error("Empty summarize")
+    // 记录总结日志
+    logger.debug(`summarize for Mr: ${summarize}`)
+    return summarize as string
+  } catch {
+    // 记录错误日志
+    logger.error(`Failed to summarize for Mr`)
+    return ""
+  }
+}
+
+export default summaryMr
--- a/controllers/manageMrEvent/utils/commenter.ts
+++ b/controllers/manageMrEvent/utils/commenter.ts
@ -0,0 +1,221 @@
+import { MergeRequestNoteSchema } from "@gitbeaker/rest"
+import { Logger } from "winston"
+
+import service from "../../../service"
+import { CreateRangeDiscussionPosition } from "../../../service/gitlab/discussions"
+import { shortenPath } from "../../../utils/pathTools"
+
+/**
+ * Gitlab评论控制器
+ */
+export class Commenter {
+  private base_sha: string
+  private head_sha: string
+  private start_sha: string
+  private project_id: number
+  private merge_request_iid: number
+  private loadingComment: MergeRequestNoteSchema | null = null
+  private comments: MergeRequestNoteSchema[] = []
+  private logger: Logger
+
+  constructor(
+    base_sha: string,
+    head_sha: string,
+    start_sha: string,
+    project_id: number,
+    merge_request_iid: number,
+    logger: Logger
+  ) {
+    this.base_sha = base_sha
+    this.head_sha = head_sha
+    this.start_sha = start_sha
+    this.project_id = project_id
+    this.merge_request_iid = merge_request_iid
+    this.logger = logger
+  }
+
+  /**
+   * 获取所有评论
+   * @returns {Promise<MergeRequestNoteSchema[]>} 返回评论列表
+   */
+  async getComments() {
+    // 如果评论列表已经存在，直接返回
+    if (this.comments.length > 0) {
+      return this.comments
+    }
+    const commentList: MergeRequestNoteSchema[] = []
+    let page = 1
+    const per_page = 100
+    while (true) {
+      // 从 GitLab 服务获取评论
+      const comments = await service.gitlab.mr.getComments(
+        this.project_id,
+        this.merge_request_iid,
+        page,
+        per_page
+      )
+      commentList.push(...comments)
+      page++
+      // 如果获取的评论数量少于每页数量，说明已经获取完毕，退出循环
+      if (comments.length < per_page) {
+        break
+      }
+    }
+    this.comments = commentList
+    return commentList
+  }
+
+  /**
+   * 创建评论
+   * @param {string} body 评论内容
+   * @returns {Promise<MergeRequestNoteSchema | undefined>} 返回创建的评论
+   */
+  async createComment(body: string) {
+    // 调用 GitLab 服务创建评论
+    const res = await service.gitlab.note.create2Mr(
+      this.project_id,
+      this.merge_request_iid,
+      body
+    )
+    if (!res) {
+      this.logger.error("Failed to create comment")
+    }
+    return res
+  }
+
+  /**
+   * 创建加载中的评论
+   * @returns {Promise<void>}
+   */
+  async createLoadingComment() {
+    const body = "小煎蛋正在处理您的合并请求，请稍等片刻。"
+    // 调用 GitLab 服务创建加载中的评论
+    const res = await service.gitlab.note.create2Mr(
+      this.project_id,
+      this.merge_request_iid,
+      body
+    )
+    if (!res) {
+      this.logger.error("Failed to create loading comment")
+    }
+    this.loadingComment = res
+  }
+
+  /**
+   * 修改加载中的评论
+   * @param {string} body 新的评论内容
+   * @returns {Promise<MergeRequestNoteSchema | undefined>} 返回修改后的评论
+   */
+  async modifyLoadingComment(body: string) {
+    if (!this.loadingComment) {
+      this.logger.error("No loading comment to modify")
+      return
+    }
+    // 调用 GitLab 服务修改加载中的评论
+    const res = await service.gitlab.note.modify2Mr(
+      this.project_id,
+      this.merge_request_iid,
+      this.loadingComment.id,
+      body
+    )
+    if (!res) {
+      this.logger.error("Failed to modify loading comment")
+    }
+    return res
+  }
+
+  /**
+   * 修改加载中的评论，如果失败则创建新的评论。
+   * @param {string} body - 评论内容。
+   * @returns {Promise<MergeRequestNoteSchema | undefined>} 返回修改或创建的评论。
+   */
+  async modifyLoadingOrCreateComment(body: string) {
+    if (this.loadingComment) {
+      const res = await this.modifyLoadingComment(body)
+      if (res) return res
+    }
+    return this.createComment(body)
+  }
+
+  /**
+   * 创建总结评论
+   * @param {string} summarizedMr 合并请求的总结
+   * @param {Map<string, string>} summarizedFileMap 文件变更的总结
+   * @returns {Promise<void>}
+   */
+  async createSummarizedComment(
+    summarizedMr: string,
+    summarizedFileMap: Map<string, string>
+  ) {
+    // 构建总结评论的内容
+    const summarizedComment = `
+# 整体摘要：
+${summarizedMr}
+
+# 文件变更：
+| 文件路径 | 变更摘要 |
+| --- | --- |
+${[...summarizedFileMap]
+  .map(([path, summary]) => `| \`${shortenPath(path)}\` | ${summary} |`)
+  .join("\n")}
+    `
+    this.logger.debug(`Summarized comment: ${summarizedComment}`)
+    // 尝试修改加载中的评论，如果失败则创建新的评论
+    const res = await this.modifyLoadingOrCreateComment(summarizedComment)
+    if (!res) {
+      this.logger.error("Failed to create summarized comment")
+    }
+  }
+
+  /**
+   * 创建审查评论
+   * @param {string} new_path 新文件路径
+   * @param {string} old_path 旧文件路径
+   * @param {number} start_line 开始行
+   * @param {number} end_line 结束行
+   * @param {string} content 评论内容
+   * @returns {Promise<void>}
+   */
+  async createReviewComment(
+    new_path: string,
+    old_path: string,
+    start_line: number,
+    end_line: number,
+    content: string
+  ) {
+    // 计算文件路径的 SHA1 哈希值
+    const hasher = new Bun.CryptoHasher("sha1")
+    hasher.update(new_path)
+    const fileSha = hasher.digest("hex")
+    // 构建评论的位置对象
+    const position: CreateRangeDiscussionPosition = {
+      base_sha: this.base_sha,
+      start_sha: this.start_sha,
+      head_sha: this.head_sha,
+      new_path,
+      old_path,
+      position_type: "text",
+      new_line: end_line,
+      line_range: {
+        start: {
+          line_code: `${fileSha}_0_${end_line}`,
+        },
+        end: {
+          line_code: `${fileSha}_0_${end_line}`,
+        },
+      },
+    }
+    // 调用 GitLab 服务创建审查评论
+    const res = await service.gitlab.discussions.create2Mr(
+      this.project_id,
+      this.merge_request_iid,
+      content,
+      position
+    )
+    if (!res) {
+      this.logger.error("Failed to create review comment")
+    }
+  }
+}
+
+export default Commenter
--- a/controllers/manageMrEvent/utils/diffTools.ts
+++ b/controllers/manageMrEvent/utils/diffTools.ts
@ -0,0 +1,342 @@
+import { CommitDiffSchema } from "@gitbeaker/rest"
+
+/**
+ * 将diff字符串拆分为多个块。
+ * @param {string | null | undefined} diff - 包含diff内容的字符串。
+ * @returns {string[]} 返回包含diff块的字符串数组。
+ */
+const splitDiff = (diff: string | null | undefined): string[] => {
+  if (diff == null) {
+    return []
+  }
+
+  const pattern = /(^@@ -(\d+),(\d+) \+(\d+),(\d+) @@).*$/gm
+
+  const result: string[] = []
+  let last = -1
+  let match: RegExpExecArray | null
+  while ((match = pattern.exec(diff)) !== null) {
+    if (last === -1) {
+      last = match.index
+    } else {
+      result.push(diff.substring(last, match.index))
+      last = match.index
+    }
+  }
+  if (last !== -1) {
+    result.push(diff.substring(last))
+  }
+  return result
+}
+
+/**
+ * 获取diff块的起始和结束行号。
+ * @param {string} diff - 包含diff内容的字符串。
+ * @returns {object | null} 返回包含旧块和新块起始和结束行号的对象，或返回null。
+ */
+const getStartEndLine = (
+  diff: string
+): {
+  oldHunk: { startLine: number; endLine: number }
+  newHunk: { startLine: number; endLine: number }
+} | null => {
+  const pattern = /(^@@ -(\d+),(\d+) \+(\d+),(\d+) @@)/gm
+  const match = pattern.exec(diff)
+  if (match != null) {
+    const oldBegin = parseInt(match[2])
+    const oldDiff = parseInt(match[3])
+    const newBegin = parseInt(match[4])
+    const newDiff = parseInt(match[5])
+    return {
+      oldHunk: {
+        startLine: oldBegin,
+        endLine: oldBegin + oldDiff - 1,
+      },
+      newHunk: {
+        startLine: newBegin,
+        endLine: newBegin + newDiff - 1,
+      },
+    }
+  } else {
+    return null
+  }
+}
+
+/**
+ * 解析diff字符串，返回旧块和新块的内容。
+ * @param {string} diff - 包含diff内容的字符串。
+ * @returns {object | null} 返回包含旧块和新块内容的对象，或返回null。
+ */
+export const parseDiff = (
+  diff: string
+): { oldHunk: string; newHunk: string } | null => {
+  const hunkInfo = getStartEndLine(diff)
+  if (hunkInfo == null) {
+    return null
+  }
+
+  const oldHunkLines: string[] = []
+  const newHunkLines: string[] = []
+
+  let newLine = hunkInfo.newHunk.startLine
+
+  const lines = diff.split("\n").slice(1) // 跳过@@行
+
+  // 如果最后一行为空，则移除
+  if (lines[lines.length - 1] === "") {
+    lines.pop()
+  }
+
+  // 跳过前3行和后3行的注释
+  const skipStart = 3
+  const skipEnd = 3
+
+  let currentLine = 0
+
+  const removalOnly = !lines.some((line) => line.startsWith("+"))
+
+  for (const line of lines) {
+    currentLine++
+    if (line.startsWith("-")) {
+      oldHunkLines.push(`${line.substring(1)}`)
+    } else if (line.startsWith("+")) {
+      newHunkLines.push(`${newLine}: ${line.substring(1)}`)
+      newLine++
+    } else {
+      // 上下文行
+      oldHunkLines.push(`${line}`)
+      if (
+        removalOnly ||
+        (currentLine > skipStart && currentLine <= lines.length - skipEnd)
+      ) {
+        newHunkLines.push(`${newLine}: ${line}`)
+      } else {
+        newHunkLines.push(`${line}`)
+      }
+      newLine++
+    }
+  }
+
+  return {
+    oldHunk: oldHunkLines.join("\n"),
+    newHunk: newHunkLines.join("\n"),
+  }
+}
+
+/**
+ * 解析文件的diff信息，返回包含起始行号、结束行号和hunk内容的数组。
+ * @param {CommitDiffSchema} file - 包含文件diff信息的对象。
+ * @returns {[number, number, string][]} 返回包含起始行号、结束行号和hunk内容的数组。
+ */
+const parseFileDiffs = (file: CommitDiffSchema): [number, number, string][] => {
+  // 获取文件的Diff，一个文件的Diff可能包含多个Hunk
+  const diffs = diffTools.splitDiff(file.diff)
+  if (diffs.length === 0) return []
+  return diffs
+    .map((diff) => {
+      const diffLines = diffTools.getStartEndLine(diff)
+      if (!diffLines) return null
+      const hunks = diffTools.parseDiff(diff)
+      if (!hunks) return null
+      const hunksStr = `
+---new_hunk---
+\`\`\`
+${hunks.newHunk}
+\`\`\`
+
+---old_hunk---
+\`\`\`
+${hunks.oldHunk}
+\`\`\`
+        `
+      return [
+        diffLines.newHunk.startLine,
+        diffLines.newHunk.endLine,
+        hunksStr,
+      ] as [number, number, string]
+    })
+    .filter((diff) => diff !== null)
+}
+
+export interface Review {
+  startLine: number
+  endLine: number
+  comment: string
+}
+
+/**
+ * 解析审查评论的函数
+ * @param {string} response - 审查评论的响应字符串
+ * @param {Array<[number, number, string]>} diffs - 差异数组，每个差异包含开始行号、结束行号和差异内容
+ * @returns {Review[]} - 返回解析后的审查评论数组
+ */
+const parseReview = (
+  response: string,
+  diffs: Array<[number, number, string]>
+): Review[] => {
+  /**
+   * 存储当前的审查评论
+   */
+  const storeReview = (): void => {
+    if (currentStartLine !== null && currentEndLine !== null) {
+      const review: Review = {
+        startLine: currentStartLine,
+        endLine: currentEndLine,
+        comment: currentComment,
+      }
+
+      let withinDiff = false
+      let bestDiffStartLine = -1
+      let bestDiffEndLine = -1
+      let maxIntersection = 0
+
+      // 查找与当前审查评论行号范围重叠最多的差异
+      for (const [startLine, endLine] of diffs) {
+        const intersectionStart = Math.max(review.startLine, startLine)
+        const intersectionEnd = Math.min(review.endLine, endLine)
+        const intersectionLength = Math.max(
+          0,
+          intersectionEnd - intersectionStart + 1
+        )
+
+        if (intersectionLength > maxIntersection) {
+          maxIntersection = intersectionLength
+          bestDiffStartLine = startLine
+          bestDiffEndLine = endLine
+          withinDiff =
+            intersectionLength === review.endLine - review.startLine + 1
+        }
+
+        if (withinDiff) break
+      }
+
+      // 如果审查评论不在任何差异范围内，进行相应处理
+      if (!withinDiff) {
+        if (bestDiffStartLine !== -1 && bestDiffEndLine !== -1) {
+          review.comment = `> 注意：此CR评论不在差异范围内，因此被映射到重叠最多的Diff。原始行号 [${review.startLine}-${review.endLine}]
+  
+  ${review.comment}`
+          review.startLine = bestDiffStartLine
+          review.endLine = bestDiffEndLine
+        } else {
+          review.comment = `> 注意：此CR评论不在差异范围内，但未找到与其重叠的Diff。原始行号 [${review.startLine}-${review.endLine}]
+  
+  ${review.comment}`
+          review.startLine = diffs[0][0]
+          review.endLine = diffs[0][1]
+        }
+      }
+
+      reviews.push(review)
+    }
+  }
+
+  /**
+   * 清理代码块中的行号
+   * @param {string} comment - 评论字符串
+   * @param {string} codeBlockLabel - 代码块标签
+   * @returns {string} - 返回清理后的评论字符串
+   */
+  const sanitizeCodeBlock = (
+    comment: string,
+    codeBlockLabel: string
+  ): string => {
+    const codeBlockStart = `\`\`\`${codeBlockLabel}`
+    const codeBlockEnd = "```"
+    const lineNumberRegex = /^ *(\d+): /gm
+
+    let codeBlockStartIndex = comment.indexOf(codeBlockStart)
+
+    while (codeBlockStartIndex !== -1) {
+      const codeBlockEndIndex = comment.indexOf(
+        codeBlockEnd,
+        codeBlockStartIndex + codeBlockStart.length
+      )
+
+      if (codeBlockEndIndex === -1) break
+
+      const codeBlock = comment.substring(
+        codeBlockStartIndex + codeBlockStart.length,
+        codeBlockEndIndex
+      )
+      const sanitizedBlock = codeBlock.replace(lineNumberRegex, "")
+
+      comment =
+        comment.slice(0, codeBlockStartIndex + codeBlockStart.length) +
+        sanitizedBlock +
+        comment.slice(codeBlockEndIndex)
+
+      codeBlockStartIndex = comment.indexOf(
+        codeBlockStart,
+        codeBlockStartIndex +
+          codeBlockStart.length +
+          sanitizedBlock.length +
+          codeBlockEnd.length
+      )
+    }
+
+    return comment
+  }
+
+  /**
+   * 清理响应字符串中的代码块
+   * @param {string} comment - 评论字符串
+   * @returns {string} - 返回清理后的评论字符串
+   */
+  const sanitizeResponse = (comment: string): string => {
+    comment = sanitizeCodeBlock(comment, "suggestion")
+    comment = sanitizeCodeBlock(comment, "diff")
+    return comment
+  }
+  const reviews: Review[] = []
+
+  // 清理响应字符串
+  response = sanitizeResponse(response.trim())
+
+  const lines = response.split("\n")
+  const lineNumberRangeRegex = /(?:^|\s)(\d+)-(\d+):\s*$/
+  const commentSeparator = "---"
+
+  let currentStartLine: number | null = null
+  let currentEndLine: number | null = null
+  let currentComment = ""
+
+  // 解析响应字符串中的每一行
+  for (const line of lines) {
+    const lineNumberRangeMatch = line.match(lineNumberRangeRegex)
+
+    if (lineNumberRangeMatch != null) {
+      storeReview()
+      currentStartLine = parseInt(lineNumberRangeMatch[1], 10)
+      currentEndLine = parseInt(lineNumberRangeMatch[2], 10)
+      currentComment = ""
+      continue
+    }
+
+    if (line.trim() === commentSeparator) {
+      storeReview()
+      currentStartLine = null
+      currentEndLine = null
+      currentComment = ""
+      continue
+    }
+
+    if (currentStartLine !== null && currentEndLine !== null) {
+      currentComment += `${line}\n`
+    }
+  }
+
+  storeReview()
+
+  return reviews
+}
+
+const diffTools = {
+  splitDiff,
+  parseDiff,
+  getStartEndLine,
+  parseFileDiffs,
+  parseReview,
+}
+
+export default diffTools
--- a/controllers/manageMrEvent/utils/inputs.ts
+++ b/controllers/manageMrEvent/utils/inputs.ts
@ -0,0 +1,128 @@
+/**
+ * 生成系统消息，包含知识截止日期和当前日期。
+ * @returns {string} 返回包含系统消息的字符串。
+ */
+const genSystemMessage = () => {
+  return `
+知识截止日期: 2023-12-01
+当前日期: ${new Date().toISOString().split("T")[0]}
+
+重要提示: 整个回复必须使用ISO代码为zh的语言
+  `
+}
+
+/**
+ * Inputs类用于存储和处理各种输入数据。
+ */
+export class Inputs {
+  systemMessage: string
+  title: string
+  description: string
+  rawSummary: string
+  shortSummary: string
+  filename: string
+  fileContent: string
+  fileDiff: string
+  patches: string
+  diff: string
+  commentChain: string
+  comment: string
+
+  /**
+   * 构造函数，初始化Inputs类的实例。
+   * @param {string} [systemMessage=""] - 系统消息。
+   * @param {string} [title="no title provided"] - 标题。
+   * @param {string} [description="no description provided"] - 描述。
+   * @param {string} [rawSummary=""] - 原始摘要。
+   * @param {string} [shortSummary=""] - 简短摘要。
+   * @param {string} [filename=""] - 文件名。
+   * @param {string} [fileContent="file contents cannot be provided"] - 文件内容。
+   * @param {string} [fileDiff="file diff cannot be provided"] - 文件差异。
+   * @param {string} [patches=""] - 补丁。
+   * @param {string} [diff="no diff"] - 差异。
+   * @param {string} [commentChain="no other comments on this patch"] - 评论链。
+   * @param {string} [comment="no comment provided"] - 评论。
+   */
+  constructor(
+    systemMessage = "",
+    title = "no title provided",
+    description = "no description provided",
+    rawSummary = "",
+    shortSummary = "",
+    filename = "",
+    fileContent = "file contents cannot be provided",
+    fileDiff = "file diff cannot be provided",
+    patches = "",
+    diff = "no diff",
+    commentChain = "no other comments on this patch",
+    comment = "no comment provided"
+  ) {
+    this.systemMessage = systemMessage || genSystemMessage()
+    this.title = title
+    this.description = description
+    this.rawSummary = rawSummary
+    this.shortSummary = shortSummary
+    this.filename = filename
+    this.fileContent = fileContent
+    this.fileDiff = fileDiff
+    this.patches = patches
+    this.diff = diff
+    this.commentChain = commentChain
+    this.comment = comment
+  }
+
+  /**
+   * 克隆当前Inputs实例。
+   * @returns {Inputs} 返回当前对象的副本。
+   */
+  clone(): Inputs {
+    return new Inputs(
+      this.systemMessage,
+      this.title,
+      this.description,
+      this.rawSummary,
+      this.shortSummary,
+      this.filename,
+      this.fileContent,
+      this.fileDiff,
+      this.patches,
+      this.diff,
+      this.commentChain,
+      this.comment
+    )
+  }
+
+  /**
+   * 渲染内容，将占位符替换为实际值。
+   * @param {string} content - 包含占位符的内容。
+   * @returns {string} 返回替换后的内容。
+   */
+  render(content: string): string {
+    if (!content) {
+      return ""
+    }
+
+    const replacements = {
+      $system_message: this.systemMessage,
+      $title: this.title,
+      $description: this.description,
+      $raw_summary: this.rawSummary,
+      $short_summary: this.shortSummary,
+      $filename: this.filename,
+      $file_content: this.fileContent,
+      $file_diff: this.fileDiff,
+      $patches: this.patches,
+      $diff: this.diff,
+      $comment_chain: this.commentChain,
+      $comment: this.comment,
+    }
+
+    for (const [key, value] of Object.entries(replacements)) {
+      if (value) {
+        content = content.replace(key, value)
+      }
+    }
+
+    return content
+  }
+}
--- a/controllers/manageMrEvent/utils/pathFilter.ts
+++ b/controllers/manageMrEvent/utils/pathFilter.ts
@ -0,0 +1,60 @@
+import { minimatch } from "minimatch"
+
+export class PathFilter {
+  // rules数组包含一组规则，每个规则是一个元组，元组的第一个元素是规则字符串，第二个元素是布尔值，表示是否排除
+  private readonly rules: Array<[string /* rule */, boolean /* exclude */]>
+
+  /**
+   * 构造函数，接受一个字符串数组作为规则
+   * @param rules - 包含路径匹配规则的字符串数组
+   */
+  constructor(rules: string[] | null = null) {
+    this.rules = []
+    if (rules != null) {
+      for (const rule of rules) {
+        const trimmed = rule?.trim() // 去除规则字符串的前后空格
+        if (trimmed) {
+          if (trimmed.startsWith("!")) {
+            // 如果规则以'!'开头，表示排除规则
+            this.rules.push([trimmed.substring(1).trim(), true])
+          } else {
+            // 否则表示包含规则
+            this.rules.push([trimmed, false])
+          }
+        }
+      }
+    }
+  }
+
+  /**
+   * 检查给定路径是否符合规则
+   * @param path - 需要检查的路径
+   * @returns boolean - 如果路径符合规则返回true，否则返回false
+   */
+  check(path: string): boolean {
+    if (this.rules.length === 0) {
+      return true // 如果没有规则，默认返回true
+    }
+
+    let included = false // 标记路径是否被包含
+    let excluded = false // 标记路径是否被排除
+    let inclusionRuleExists = false // 标记是否存在包含规则
+
+    for (const [rule, exclude] of this.rules) {
+      if (minimatch(path, rule)) {
+        // 使用minimatch库检查路径是否匹配规则
+        if (exclude) {
+          excluded = true // 如果匹配排除规则，设置excluded为true
+        } else {
+          included = true // 如果匹配包含规则，设置included为true
+        }
+      }
+      if (!exclude) {
+        inclusionRuleExists = true // 如果存在包含规则，设置inclusionRuleExists为true
+      }
+    }
+
+    // 返回值逻辑：如果不存在包含规则或路径被包含且未被排除，则返回true
+    return (!inclusionRuleExists || included) && !excluded
+  }
+}
--- a/controllers/manageMrEvent/utils/prompts.ts
+++ b/controllers/manageMrEvent/utils/prompts.ts
@ -0,0 +1,206 @@
+import { type Inputs } from "./inputs"
+
+export class Prompts {
+  // 总结单文件修改内容
+  summarizeFileDiff = `
+$system_message
+
+## Gitlab 合并请求 标题
+
+\`\`\`
+$title
+\`\`\`
+
+## 描述
+
+\`\`\`
+$description
+\`\`\`
+
+## 差异
+
+\`\`\`diff
+$file_diff
+\`\`\`
+
+## 说明
+
+我希望你能简洁地总结这个差异，不超过100个字。
+如果适用，你的总结应包括对导出函数、全局数据结构和变量签名的更改的说明，以及任何可能影响代码外部接口或行为的更改。
+不要返回任何标题，直接回复差异总结。
+`
+  // 文件是否需要Review
+  triageFileDiff = `在总结下方，我还希望你根据以下标准将差异分类为\`NEEDS_REVIEW\`或\`APPROVED\`：
+
+- 如果差异涉及任何逻辑或功能的修改，即使它们看起来很小，也将其分类为\`NEEDS_REVIEW\`。这包括对控制结构、函数调用或变量赋值的更改，这些更改可能会影响代码的行为。
+- 如果差异仅包含不影响代码逻辑的非常小的更改，例如修正拼写错误、格式或为了清晰起见重命名变量、修改引入顺序，将其分类为\`APPROVED\`。
+
+请彻底评估差异，并考虑更改行数、对整个系统的潜在影响以及引入新错误或安全漏洞的可能性。
+在有疑问时，总是倾向于谨慎并将差异分类为\`NEEDS_REVIEW\`。
+
+你必须严格按照以下格式对差异进行分类：
+[TRIAGE]: <NEEDS_REVIEW or APPROVED>
+
+重要提示：
+- 返回值不要包含\`\`\`。
+- 在你的总结中不要提到文件需要彻底审核或警告潜在问题。
+- 不要提供任何将差异分类为\`NEEDS_REVIEW\`或\`APPROVED\`的理由。
+- 在总结中不要提到这些更改会影响代码的逻辑或功能。你只能使用上述分类格式来指示。
+`
+  // 总结前缀
+  summarizePrefix = `以下是你为文件生成的更改摘要：
+      \`\`\`
+      $raw_summary
+      \`\`\`
+
+`
+  // 总结
+  summarize = `你的任务是提供一个简明的合并请求的更改摘要。此摘要将在代码审查时用作提示，必须非常清晰以便AI机器人理解。
+
+说明：
+
+- 返回值不要包含\`\`\`。
+- 仅关注总结合并请求中的更改，并坚持事实。
+- 不要向机器人提供任何关于如何执行审查的说明。
+- 不要提到文件需要彻底审核或警告潜在问题。
+- 不要提到这些更改会影响代码的逻辑或功能。
+- 不要提到任何单独的文件名，在代码审查中这部分会另外提供。
+- 返回纯文本内容，叙述整体的修改，不得超过500字。
+`
+  // 审查文件差异
+  reviewFileDiff = `# Gitlab 合并请求
+## 标题
+
+\`$title\` 
+
+## 描述
+
+\`\`\`
+$description
+\`\`\`
+
+## 更改摘要
+
+\`\`\`
+$short_summary
+\`\`\`
+
+## 重要说明
+你是一个出类拔萃的CodeReviewer，现在是时候展示你的技能了！
+
+输入：带有行号的新hunks和旧hunks（替换的代码）。hunks代表不完整的代码片段。
+
+附加上下文：合并请求的标题、描述、摘要和评论链。
+
+任务：使用提供的上下文审查新hunks中的实质性问题，并在必要时回复评论。
+
+输出：在markdown中使用确切的行号范围进行审查评论。单行评论的起始和结束行号必须相同。必须使用以下示例响应格式：
+使用带有相关语言标识符的围栏代码块（fenced code blocks）。
+不要用行号注释代码片段。
+正确格式化和缩进代码。
+不要使用\`suggestion\`代码块。
+对于问题修复：
+- 使用\`diff\`代码块，用\`+\`或\`-\`标记更改。带有修复片段的评论的行号范围必须与新hunk中要替换的范围完全匹配。
+- 不要提供一般反馈、更改摘要、变更解释或对良好代码的赞扬。
+- 更多的关注于逻辑错误、边界情况和性能问题。
+- 忽略代码风格问题、引入顺序问题及输出打印等不影响代码功能的问题，除非它们导致错误。
+- 仅专注于根据给定的上下文提供具体、客观的见解，避免对系统潜在影响的广泛评论或质疑更改意图。
+- 如果一个区块反应了多个问题，请一起回复。
+
+如果在行范围内没有发现问题，只在审查部分回复纯文本 \`LGTM!\` 。
+且如果返回\`LGTM!\`的行号连续，则可以合并为一个区块。
+
+## 示例
+
+### 示例更改
+
+---new_hunk---
+\`\`\`
+  z = x / y
+    return z
+
+20: def add(x, y):
+21:     z = x + y
+22:     retrn z
+23: 
+24: def multiply(x, y):
+25:     return x * y
+
+def subtract(x, y):
+  z = x - y
+\`\`\`
+  
+---old_hunk---
+\`\`\`
+  z = x / y
+    return z
+
+def add(x, y):
+    return x + y
+
+def subtract(x, y):
+    z = x - y
+\`\`\`
+
+---comment_chains---
+\`\`\`
+请审查此更改。
+\`\`\`
+
+---end_change_section---
+
+### 示例响应
+
+22-22:
+add函数中有一个语法错误。
+\`\`\`diff
+-    retrn z
+    return z
+\`\`\`
+---
+24-25:
+LGTM!
+---
+
+## 你要审查的\`$filename\`的更改
+
+$patches
+
+## 审查结果
+`
+
+  constructor() {}
+
+  /**
+   * 文件差异摘要。
+   * @param {Inputs} inputs - 输入对象。
+   * @param {boolean} needReview - 是否需要审查。
+   * @returns {string} 渲染后的字符串。
+   */
+  renderSummarizeFileDiff(inputs: Inputs, needReview: boolean): string {
+    let prompt = this.summarizeFileDiff
+    if (needReview) {
+      prompt += this.triageFileDiff
+    }
+    return inputs.render(prompt)
+  }
+
+  /**
+   * 摘要。
+   * @param {Inputs} inputs - 输入对象。
+   * @returns {string} 渲染后的字符串。
+   */
+  renderSummarize(inputs: Inputs): string {
+    const prompt = this.summarizePrefix + this.summarize
+    return inputs.render(prompt)
+  }
+
+  /**
+   * 文件差异审查。
+   * @param {Inputs} inputs - 输入对象。
+   * @returns {string} 渲染后的字符串。
+   */
+  renderReviewFileDiff(inputs: Inputs): string {
+    return inputs.render(this.reviewFileDiff)
+  }
+}
--- a/index.ts
+++ b/index.ts
@ -1,3 +1,6 @@
+import { v4 as uuid } from "uuid"
+
+import loggerIns from "./log"
 import { manageCIMonitorReq } from "./routes/ci"
 import { manageGitlabEventReq } from "./routes/event"
 import initSchedule from "./schedule"
@ -12,14 +15,16 @@ const PREFIX = "/gitlab_monitor"
 const server = Bun.serve({
  async fetch(req) {
    try {
+      // 添加请求ID
+      const logger = loggerIns.child({ requestId: uuid() })
      // 路由处理
      const { exactCheck, fullCheck } = makeCheckPathTool(req.url, PREFIX)
      // 非根路由打印
-      if (!fullCheck("/")) console.log("🚀 ~ serve ~ req.url", req.url)
+      if (!fullCheck("/")) logger.info(`${req.method} ${req.url}`)
      // CI 监控
      if (exactCheck("/ci")) return manageCIMonitorReq(req)
      // Gitlab 事件
-      if (exactCheck("/event")) return manageGitlabEventReq(req)
+      if (exactCheck("/event")) return manageGitlabEventReq(req, logger)
      // 其他
      return netTool.ok("hello, there is gitlab monitor, glade to serve you!")
    } catch (error: any) {
@ -31,4 +36,4 @@ const server = Bun.serve({
  port: 3000,
 })

-console.log(`Listening on ${server.hostname}:${server.port}`)
+loggerIns.info(`Listening on ${server.hostname}:${server.port}`)
--- a/log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json
+++ b/log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json
@ -0,0 +1,15 @@
+{
+    "keep": {
+        "days": true,
+        "amount": 14
+    },
+    "auditLog": "log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json",
+    "files": [
+        {
+            "date": 1723259039565,
+            "name": "log/application-2024-08-10.log",
+            "hash": "fdd4f67d11fe79d1aedf242b8d2e1894fbd6a4679331098818c2324ec81678cd"
+        }
+    ],
+    "hashType": "sha256"
+}
--- a/log/index.ts
+++ b/log/index.ts
@ -0,0 +1,38 @@
+import "winston-daily-rotate-file"
+
+import winston, { format } from "winston"
+
+const isProd = process.env.NODE_ENV === "production"
+
+const dailyRotateFileTransport = new winston.transports.DailyRotateFile({
+  filename: "./log/application-%DATE%.log",
+  datePattern: "YYYY-MM-DD",
+  zippedArchive: true,
+  maxSize: "20m",
+  maxFiles: "14d",
+})
+
+const transports: any[] = [new winston.transports.Console()]
+if (isProd) {
+  transports.push(dailyRotateFileTransport)
+}
+
+const loggerIns = winston.createLogger({
+  level: "info",
+  format: format.combine(
+    format.colorize({
+      level: !isProd,
+    }), // 开发环境下输出彩色日志
+    format.simple(), // 简单文本格式化
+    format.timestamp({ format: "YYYY-MM-DD HH:mm:ss" }),
+    format.printf(({ level, message, timestamp, requestId }) => {
+      const singleLineMessage = isProd
+        ? message.replace(/\n/g, " ") // 将换行符替换为空格
+        : message
+      return `${timestamp} [${level}] ${requestId ? `[RequestId: ${requestId}]` : ""}: ${singleLineMessage}`
+    })
+  ),
+  transports,
+})
+
+export default loggerIns
--- a/package.json
+++ b/package.json
@ -19,6 +19,7 @@
    "@commitlint/cli": "^19.3.0",
    "@commitlint/config-conventional": "^19.2.2",
    "@eslint/js": "^9.7.0",
+    "@gitbeaker/rest": "^40.1.2",
    "@types/lodash": "^4.14.202",
    "@types/node-schedule": "^2.1.6",
    "bun-types": "latest",
@ -35,10 +36,16 @@
    "typescript": "^5.0.0"
  },
  "dependencies": {
+    "@dqbd/tiktoken": "^1.0.15",
+    "@langchain/openai": "^0.2.6",
    "lodash": "^4.17.21",
+    "minimatch": "^10.0.1",
    "moment": "^2.30.1",
    "node-schedule": "^2.1.1",
    "p-limit": "^6.1.0",
-    "pocketbase": "^0.21.1"
+    "pocketbase": "^0.21.1",
+    "uuid": "^10.0.0",
+    "winston": "^3.14.1",
+    "winston-daily-rotate-file": "^5.0.0"
  }
 }
--- a/routes/event/index.ts
+++ b/routes/event/index.ts
@ -1,3 +1,6 @@
+import { Logger } from "winston"
+
+import manageMrEvent from "../../controllers/manageMrEvent"
 import managePipelineEvent from "../../controllers/managePipelineEvent"
 import netTool from "../../service/netTool"
 import { Gitlab } from "../../types/gitlab"
@ -7,15 +10,26 @@ import { Gitlab } from "../../types/gitlab"
 * @param {Request} req - 请求对象。
 * @returns {Promise<Response>} - 响应对象。
 */
-export const manageGitlabEventReq = async (req: Request): Promise<Response> => {
+export const manageGitlabEventReq = async (
+  req: Request,
+  logger: Logger
+): Promise<Response> => {
  const apiKey = req.headers.get("x-gitlab-token")
+  logger.info(`x-gitlab-token: ${apiKey}`)
  if (!apiKey) return netTool.badRequest("x-gitlab-token is required!")
  const eventType = req.headers.get("x-gitlab-event")
-  // 只处理流水线钩子
+  logger.info(`x-gitlab-event: ${eventType}`)
+  // 处理流水线钩子
  if (eventType === "Pipeline Hook") {
    const body = (await req.json()) as Gitlab.PipelineEvent
    const params = new URLSearchParams(req.url.split("?")[1])
    return managePipelineEvent.manageRawEvent(body, apiKey, params)
  }
+  // 处理MR钩子
+  if (eventType === "Merge Request Hook") {
+    const body = (await req.json()) as Gitlab.MergeRequestEvent
+    logger.debug(`body: ${JSON.stringify(body)}`)
+    return manageMrEvent.manageRawEvent(body, logger)
+  }
  return netTool.ok()
 }
--- a/script/mr/changes.json
+++ b/script/mr/changes.json
--- a/script/mr/comments.json
+++ b/script/mr/comments.json
@ -0,0 +1,443 @@
+[
+  {
+    "id": 9591828,
+    "type": null,
+    "body": "mentioned in commit 17a35cfb47b46e85f507c5f602907ffc429d40c0",
+    "attachment": null,
+    "author": {
+      "id": 30382,
+      "username": "wuting7",
+      "name": "吴婷",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/30382/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/wuting7"
+    },
+    "created_at": "2024-08-09T10:54:28.295+08:00",
+    "updated_at": "2024-08-09T10:54:28.297+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591827,
+    "type": null,
+    "body": "approved this merge request",
+    "attachment": null,
+    "author": {
+      "id": 30382,
+      "username": "wuting7",
+      "name": "吴婷",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/30382/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/wuting7"
+    },
+    "created_at": "2024-08-09T10:54:25.753+08:00",
+    "updated_at": "2024-08-09T10:54:25.755+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591699,
+    "type": null,
+    "body": "resolved all threads",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:49:03.570+08:00",
+    "updated_at": "2024-08-09T10:49:03.577+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591698,
+    "type": "DiffNote",
+    "body": "删完了捏",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:49:03.134+08:00",
+    "updated_at": "2024-08-09T10:49:03.134+08:00",
+    "system": false,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "commit_id": null,
+    "position": {
+      "base_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "start_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "head_sha": "4e7ef3dfd9a731f265c0b7e65d96476ad35e1643",
+      "old_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "new_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "position_type": "text",
+      "old_line": null,
+      "new_line": 179,
+      "line_range": {
+        "start": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        },
+        "end": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        }
+      }
+    },
+    "resolvable": true,
+    "resolved": true,
+    "resolved_by": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "resolved_at": "2024-08-09T10:49:03.545+08:00",
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591676,
+    "type": null,
+    "body": "added 1 commit\n\n<ul><li>9af87594 - chore: 删除示例数据</li></ul>\n\n[Compare with previous version](/cloudml-visuals/fe/cloud-ml-fe/-/merge_requests/477/diffs?diff_id=6505933&start_sha=4e7ef3dfd9a731f265c0b7e65d96476ad35e1643)",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:48:17.487+08:00",
+    "updated_at": "2024-08-09T10:48:17.489+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591675,
+    "type": "DiffNote",
+    "body": "changed this line in [version 2 of the diff](/cloudml-visuals/fe/cloud-ml-fe/-/merge_requests/477/diffs?diff_id=6505933&start_sha=4e7ef3dfd9a731f265c0b7e65d96476ad35e1643#cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_179_160)",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:48:17.244+08:00",
+    "updated_at": "2024-08-09T10:48:17.244+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "commit_id": null,
+    "position": {
+      "base_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "start_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "head_sha": "4e7ef3dfd9a731f265c0b7e65d96476ad35e1643",
+      "old_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "new_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "position_type": "text",
+      "old_line": null,
+      "new_line": 179,
+      "line_range": {
+        "start": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        },
+        "end": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        }
+      }
+    },
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591335,
+    "type": "DiffNote",
+    "body": "算了，我改到example里吧",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:34:10.533+08:00",
+    "updated_at": "2024-08-09T10:34:10.533+08:00",
+    "system": false,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "commit_id": null,
+    "position": {
+      "base_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "start_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "head_sha": "4e7ef3dfd9a731f265c0b7e65d96476ad35e1643",
+      "old_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "new_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "position_type": "text",
+      "old_line": null,
+      "new_line": 179,
+      "line_range": {
+        "start": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        },
+        "end": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        }
+      }
+    },
+    "resolvable": true,
+    "resolved": true,
+    "resolved_by": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "resolved_at": "2024-08-09T10:49:03.545+08:00",
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591318,
+    "type": null,
+    "body": "resolved all threads",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:33:43.448+08:00",
+    "updated_at": "2024-08-09T10:33:43.454+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9591317,
+    "type": "DiffNote",
+    "body": "只是删了引用，留点原始数据好看格式",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-09T10:33:42.907+08:00",
+    "updated_at": "2024-08-09T10:33:42.907+08:00",
+    "system": false,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "commit_id": null,
+    "position": {
+      "base_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "start_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "head_sha": "4e7ef3dfd9a731f265c0b7e65d96476ad35e1643",
+      "old_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "new_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "position_type": "text",
+      "old_line": null,
+      "new_line": 179,
+      "line_range": {
+        "start": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        },
+        "end": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        }
+      }
+    },
+    "resolvable": true,
+    "resolved": true,
+    "resolved_by": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "resolved_at": "2024-08-09T10:49:03.545+08:00",
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9581569,
+    "type": "DiffNote",
+    "body": "mock 数据不是都删了麽？",
+    "attachment": null,
+    "author": {
+      "id": 30382,
+      "username": "wuting7",
+      "name": "吴婷",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/30382/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/wuting7"
+    },
+    "created_at": "2024-08-08T15:20:24.391+08:00",
+    "updated_at": "2024-08-08T15:20:24.391+08:00",
+    "system": false,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "commit_id": null,
+    "position": {
+      "base_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "start_sha": "e65a68d698f0e731285b4fbdee34102a21442803",
+      "head_sha": "4e7ef3dfd9a731f265c0b7e65d96476ad35e1643",
+      "old_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "new_path": "src/pages/AIWorkbench/typings/workflowIns.ts",
+      "position_type": "text",
+      "old_line": null,
+      "new_line": 179,
+      "line_range": {
+        "start": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        },
+        "end": {
+          "line_code": "cf42380df1ea3d449fda4ef0d59a9e3a2d02aeb3_174_179",
+          "type": "new",
+          "old_line": null,
+          "new_line": 179
+        }
+      }
+    },
+    "resolvable": true,
+    "resolved": true,
+    "resolved_by": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "resolved_at": "2024-08-09T10:49:03.545+08:00",
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9581351,
+    "type": null,
+    "body": "assigned to @zhaoyingbo",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-08T15:11:28.865+08:00",
+    "updated_at": "2024-08-08T15:11:28.891+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  },
+  {
+    "id": 9581349,
+    "type": null,
+    "body": "requested review from @wuting7",
+    "attachment": null,
+    "author": {
+      "id": 10011,
+      "username": "zhaoyingbo",
+      "name": "赵英博",
+      "state": "active",
+      "avatar_url": "https://git.n.xiaomi.com/uploads/-/system/user/avatar/10011/avatar.png",
+      "web_url": "https://git.n.xiaomi.com/zhaoyingbo"
+    },
+    "created_at": "2024-08-08T15:11:28.338+08:00",
+    "updated_at": "2024-08-08T15:11:28.340+08:00",
+    "system": true,
+    "noteable_id": 2351580,
+    "noteable_type": "MergeRequest",
+    "resolvable": false,
+    "confidential": false,
+    "noteable_iid": 477,
+    "commands_changes": {}
+  }
+]
--- a/script/mr/event.json
+++ b/script/mr/event.json
@ -0,0 +1,17 @@
+{
+  "object_kind": "merge_request",
+  "event_type": "merge_request",
+  "project": {
+    "id": 139032
+  },
+  "object_attributes": {
+    "iid": 490,
+    "state": "merged",
+    "changes": {
+      "updated_at": {
+        "previous": "2021-07-01 10:00:00 UTC",
+        "current": "2021-07-01 10:30:00 UTC"
+      }
+    }
+  }
+}
--- a/service/gitlab/discussions.ts
+++ b/service/gitlab/discussions.ts
@ -0,0 +1,64 @@
+import { DiscussionSchema } from "@gitbeaker/rest"
+
+import netTool from "../netTool"
+import { GITLAB_AUTH_HEADER, GITLAB_BASE_URL, gitlabReqWarp } from "./tools"
+
+export interface CreateRangeDiscussionPosition {
+  base_sha: string
+  start_sha: string
+  head_sha: string
+  new_path: string
+  old_path: string
+  position_type: "text"
+  new_line: number
+  line_range: {
+    start: {
+      line_code: string
+    }
+    end: {
+      line_code: string
+    }
+  }
+}
+
+/**
+ * 获取合并请求的讨论列表。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @returns {Promise<DiscussionSchema[]>} 返回包含讨论列表的Promise。
+ */
+const getList = async (project_id: number, merge_request_iid: number) => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/discussions`
+  return gitlabReqWarp<DiscussionSchema[]>(
+    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
+    []
+  )
+}
+
+/**
+ * 创建合并请求的讨论。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @param {string} body - 讨论内容。
+ * @param {CreateRangeDiscussionPosition} position - 讨论位置。
+ * @returns {Promise<DiscussionSchema | null>} 返回包含创建的讨论的Promise。
+ */
+const create2Mr = async (
+  project_id: number,
+  merge_request_iid: number,
+  body: string,
+  position: CreateRangeDiscussionPosition
+) => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/discussions`
+  return gitlabReqWarp<DiscussionSchema>(
+    () => netTool.post(URL, { body, position }, {}, GITLAB_AUTH_HEADER),
+    null
+  )
+}
+
+const discussions = {
+  getList,
+  create2Mr,
+}
+
+export default discussions
--- a/service/gitlab/index.ts
+++ b/service/gitlab/index.ts
@ -1,5 +1,8 @@
 import badge from "./badge"
 import commit from "./commit"
+import discussions from "./discussions"
+import mr from "./mr"
+import note from "./note"
 import pipeline from "./pipeline"
 import project from "./project"

@ -8,6 +11,9 @@ const gitlab = {
  badge,
  commit,
  pipeline,
+  mr,
+  note,
+  discussions,
 }

 export default gitlab
--- a/service/gitlab/mr.ts
+++ b/service/gitlab/mr.ts
@ -1,20 +1,84 @@
+import type {
+  ExpandedMergeRequestSchema,
+  MergeRequestChangesSchema,
+  MergeRequestDiffVersionsSchema,
+  MergeRequestNoteSchema,
+} from "@gitbeaker/rest"
+
 import netTool from "../netTool"
 import { GITLAB_AUTH_HEADER, GITLAB_BASE_URL, gitlabReqWarp } from "./tools"

-const getDiffs = async (project_id: number, merge_request_iid: number) => {
+/**
+ * 获取合并请求的评论，支持分页。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @param {number} [page=1] - 页码。
+ * @param {number} [per_page=20] - 每页的评论数。
+ * @returns {Promise<MergeRequestNoteSchema[]>} 返回包含评论的Promise。
+ */
+const getComments = async (
+  project_id: number,
+  merge_request_iid: number,
+  page: number = 1,
+  per_page: number = 20
+): Promise<MergeRequestNoteSchema[]> => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/notes?page=${page}&per_page=${per_page}`
+  return gitlabReqWarp<MergeRequestNoteSchema[]>(
+    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
+    []
+  )
+}
+
+/**
+ * 获取合并请求的变更。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @returns {Promise<MergeRequestChangesSchema | null>} 返回包含变更的Promise。
+ */
+const getChanges = async (project_id: number, merge_request_iid: number) => {
  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/changes`
-  const res = await gitlabReqWarp(
+  return gitlabReqWarp<MergeRequestChangesSchema | null>(
    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
    null
  )
-  if (res === null) return null
-  return res
+}
+
+/**
+ * 获取合并请求的详细信息。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @returns {Promise<ExpandedMergeRequestSchema | null>} 返回包含详细信息的Promise。
+ */
+const getDetail = async (project_id: number, merge_request_iid: number) => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}`
+  return gitlabReqWarp<ExpandedMergeRequestSchema | null>(
+    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
+    null
+  )
+}
+
+/**
+ * 获取合并请求的差异版本。
+ * @param {number} project_id - 项目ID。
+ * @param {number} merge_request_iid - 合并请求IID。
+ * @returns {Promise<any[]>} 返回包含差异版本的Promise。
+ */
+const getDiffVersions = async (
+  project_id: number,
+  merge_request_iid: number
+) => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/versions`
+  return gitlabReqWarp<MergeRequestDiffVersionsSchema[]>(
+    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
+    []
+  )
 }

 const mr = {
-  getDiffs,
+  getChanges,
+  getDetail,
+  getComments,
+  getDiffVersions,
 }

 export default mr
-
-getDiffs(139032, 484).then(console.log)
--- a/service/gitlab/note.ts
+++ b/service/gitlab/note.ts
@ -0,0 +1,51 @@
+import { MergeRequestNoteSchema } from "@gitbeaker/rest"
+
+import netTool from "../netTool"
+import { GITLAB_AUTH_HEADER, GITLAB_BASE_URL, gitlabReqWarp } from "./tools"
+
+/**
+ * 创建一个新的合并请求备注
+ * @param {number} project_id - 项目ID
+ * @param {number} merge_request_iid - 合并请求IID
+ * @param {string} body - 备注内容
+ * @returns {Promise<MergeRequestNoteSchema>} - 返回包含新创建备注的Promise
+ */
+const create2Mr = async (
+  project_id: number,
+  merge_request_iid: number,
+  body: string
+): Promise<MergeRequestNoteSchema> => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/notes`
+  return gitlabReqWarp<MergeRequestNoteSchema>(
+    () => netTool.post(URL, { body }, {}, GITLAB_AUTH_HEADER),
+    null
+  )
+}
+
+/**
+ * 修改一个现有的合并请求备注
+ * @param {number} project_id - 项目ID
+ * @param {number} merge_request_iid - 合并请求IID
+ * @param {number} note_id - 备注ID
+ * @param {string} body - 新的备注内容
+ * @returns {Promise<MergeRequestNoteSchema>} - 返回包含修改后备注的Promise
+ */
+const modify2Mr = async (
+  project_id: number,
+  merge_request_iid: number,
+  note_id: number,
+  body: string
+): Promise<MergeRequestNoteSchema> => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/merge_requests/${merge_request_iid}/notes/${note_id}`
+  return gitlabReqWarp<MergeRequestNoteSchema>(
+    () => netTool.put(URL, { body }, {}, GITLAB_AUTH_HEADER),
+    null
+  )
+}
+
+const note = {
+  create2Mr,
+  modify2Mr,
+}
+
+export default note
--- a/service/gitlab/repository.ts
+++ b/service/gitlab/repository.ts
@ -0,0 +1,27 @@
+import netTool from "../netTool"
+import { GITLAB_AUTH_HEADER, GITLAB_BASE_URL, gitlabReqWarp } from "./tools"
+
+/**
+ * 获取指定项目中某个文件的内容。
+ * @param {number} project_id - 项目ID。
+ * @param {string} path - 文件路径。
+ * @param {string} ref - 分支或标签名称。
+ * @returns {Promise<string>} 返回包含文件内容的Promise。
+ */
+const getFileContent = async (
+  project_id: number,
+  path: string,
+  ref: string
+) => {
+  const URL = `${GITLAB_BASE_URL}/projects/${project_id}/repository/files/${encodeURIComponent(path)}/raw?ref=${ref}`
+  return gitlabReqWarp<string>(
+    () => netTool.get(URL, {}, GITLAB_AUTH_HEADER),
+    ""
+  )
+}
+
+const repository = {
+  getFileContent,
+}
+
+export default repository
--- a/service/netTool.ts
+++ b/service/netTool.ts
@ -33,7 +33,7 @@ const logResponse = (
    requestBody,
    responseBody,
  }
-  console.log("🚀 ~ responseLog:", JSON.stringify(responseLog, null, 2))
+  // console.log("🚀 ~ responseLog:", JSON.stringify(responseLog, null, 2))
  return responseLog
 }

--- a/test/log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json
+++ b/test/log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json
@ -0,0 +1,15 @@
+{
+    "keep": {
+        "days": true,
+        "amount": 14
+    },
+    "auditLog": "log/.477ef71694ce4c791e6f91cf40117d4a85785c6b-audit.json",
+    "files": [
+        {
+            "date": 1723423178431,
+            "name": "log/application-2024-08-12.log",
+            "hash": "0bc427a58e73e382ac12bffa78e2d17ab09717e51a940fe5d71f786f5f1e6bcf"
+        }
+    ],
+    "hashType": "sha256"
+}
--- a/test/manageGitlabEventReq.test.ts
+++ b/test/manageGitlabEventReq.test.ts
@ -1,5 +1,6 @@
 import { expect, test } from "bun:test"

+import loggerIns from "../log"
 import { manageGitlabEventReq } from "../routes/event"
 import netTool from "../service/netTool"
 import { Gitlab } from "../types/gitlab"
@ -59,6 +60,8 @@ test("manageGitlabEventReq", async () => {
    }
  )

-  const res = await manageGitlabEventReq(req)
+  const logger = loggerIns.child({ name: "test" })
+
+  const res = await manageGitlabEventReq(req, logger)
  expect(res).toEqual(netTool.ok())
 })
--- a/test/manageMrEvent.test.ts
+++ b/test/manageMrEvent.test.ts
@ -0,0 +1,38 @@
+import { expect, test } from "bun:test"
+
+import loggerIns from "../log"
+import { manageGitlabEventReq } from "../routes/event"
+import netTool from "../service/netTool"
+import { Gitlab } from "../types/gitlab"
+
+test("manageMrEvent", async () => {
+  const headers = new Headers({
+    "x-gitlab-token": "uwnpzb9hvoft28h",
+    "x-gitlab-event": "Merge Request Hook",
+  })
+
+  const body: Gitlab.MergeRequestEvent = {
+    object_kind: "merge_request",
+    event_type: "merge_request",
+    project: {
+      id: 139032,
+    },
+    object_attributes: {
+      iid: 502,
+      state: "opened",
+    },
+  }
+
+  const req = new Request(
+    "https://lark-egg.ai.xiaomi.com/gitlab_monitor/event",
+    {
+      method: "POST",
+      headers: headers,
+      body: JSON.stringify(body),
+    }
+  )
+
+  const logger = loggerIns.child({ requestId: "test" })
+  const res = await manageGitlabEventReq(req, logger)
+  expect(res).toEqual(netTool.ok())
+}, 100000)
--- a/test/parseReview.test.ts
+++ b/test/parseReview.test.ts
@ -0,0 +1,32 @@
+import { test } from "bun:test"
+
+import diffTools from "../controllers/manageMrEvent/utils/diffTools"
+import loggerIns from "../log"
+
+test("parseReview", async () => {
+  const logger = loggerIns.child({ requestId: "test" })
+  const response = `
+  1-12:
+LGTM!
+---
+67-70:
+在生产环境中，\`console.log\`可能会导致性能问题，建议移除或使用更合适的方式进行调试。
+\`\`\`diff
+-                     console.log(
+-                       '🚀 ~ file: index.tsx:68 ~ ModelSquare ~ item.apiDoc:',
+-                       item.apiDoc,
+-                     );
+\`\`\`
+---
+`
+  const diffs = [
+    [
+      1,
+      12,
+      "\\n---new_hunk---\\n```\\n import { Button, Space, Tag, Typography } from 'antd';\\n import classnames from 'classnames';\\n import { memo } from 'react';\\n4: \\n5:  import ChatGLM from '../../assets/ChatGLM.png';\\n6: import CVLM from '../../assets/CVLM.png';\\n7:  import inner from '../../assets/inner.png';\\n8:  import Llama from '../../assets/Llama.png';\\n9: import MICV from '../../assets/MICV.png';\\n import MiniGPT from '../../assets/MiniGPT.png';\\n import Mixtral from '../../assets/Mixtral.png';\\n import Qwen from '../../assets/Qwen.png';\\n```\\n\\n---old_hunk---\\n```\\n import { Button, Space, Tag, Typography } from 'antd';\\n import classnames from 'classnames';\\n import { memo } from 'react';\\n import ChatGLM from '../../assets/ChatGLM.png';\\n import inner from '../../assets/inner.png';\\n import Llama from '../../assets/Llama.png';\\n import MiniGPT from '../../assets/MiniGPT.png';\\n import Mixtral from '../../assets/Mixtral.png';\\n import Qwen from '../../assets/Qwen.png';\\n```\\n        ",
+    ],
+  ] as Array<[number, number, string]>
+
+  const res = diffTools.parseReview(response, diffs)
+  logger.info(JSON.stringify(res))
+}, 100000)
--- a/test/pathFilter.test.ts
+++ b/test/pathFilter.test.ts
@ -0,0 +1,28 @@
+import { expect, test } from "bun:test"
+
+import { PathFilter } from "../controllers/manageMrEvent/utils/pathFilter"
+
+// 测试用例
+test("PathFilter check method", () => {
+  // 示例规则数组
+  const rules = [
+    "*.js", // 包含所有.js文件
+    "!*.test.js", // 排除所有.test.js文件
+    "src/**", // 包含src目录下的所有文件
+    "!src/tmp/**", // 排除src/tmp目录下的所有文件
+  ]
+
+  // 创建PathFilter实例
+  const pathFilter = new PathFilter(rules)
+  // 测试包含.js文件
+  expect(pathFilter.check("example.js")).toBe(true)
+
+  // 测试排除.test.js文件
+  expect(pathFilter.check("example.test.js")).toBe(false)
+
+  // 测试包含src目录下的文件
+  expect(pathFilter.check("src/index.js")).toBe(true)
+
+  // 测试排除src/tmp目录下的文件
+  expect(pathFilter.check("src/tmp/temp.js")).toBe(false)
+})
--- a/test/service/discussion.test.ts
+++ b/test/service/discussion.test.ts
@ -0,0 +1,43 @@
+import { test } from "bun:test"
+
+import service from "../../service"
+
+// 测试用例
+test("Gitlab Discussion", async () => {
+  const project_id = 139032
+  const merge_request_iid = 488
+
+  const res = await service.gitlab.discussions.getList(
+    project_id,
+    merge_request_iid
+  )
+  console.log(res)
+
+  const body = "LGTM!"
+  const position: any = {
+    base_sha: "17a35cfb47b46e85f507c5f602907ffc429d40c0",
+    start_sha: "2d6f9d6d5cbfd323e06e37859ee9b19c3578619c",
+    head_sha: "710458e37476915e585d480961786e7eedbc79fb",
+    old_path: "src/pages/AIWorkbench/controller/index.ts",
+    new_path: "src/pages/AIWorkbench/controller/index.ts",
+    position_type: "text",
+    new_line: 21,
+    line_range: {
+      start: {
+        line_code: "6988b6e297ca054b9ffaa69f4a0f4a954e749f78_0_21",
+      },
+      end: {
+        line_code: "6988b6e297ca054b9ffaa69f4a0f4a954e749f78_0_25",
+      },
+    },
+  }
+
+  const res2 = await service.gitlab.discussions.create2Mr(
+    project_id,
+    merge_request_iid,
+    body,
+    position
+  )
+
+  console.log(res2)
+}, 10000)
--- a/test/service/mr.test.ts
+++ b/test/service/mr.test.ts
@ -0,0 +1,15 @@
+import { test } from "bun:test"
+
+import service from "../../service"
+
+// 测试用例
+test("Gitlab MR", async () => {
+  const project_id = 139032
+  const merge_request_iid = 4889
+
+  const res = await service.gitlab.mr.getDiffVersions(
+    project_id,
+    merge_request_iid
+  )
+  console.log(res)
+}, 10000)
--- a/test/service/note.test.ts
+++ b/test/service/note.test.ts
@ -0,0 +1,41 @@
+import { test } from "bun:test"
+
+import service from "../../service"
+
+const sleep = (ms: number) => new Promise((resolve) => setTimeout(resolve, ms))
+
+// 测试用例
+test("Gitlab Note", async () => {
+  const summarizedMr = `增加了两个新的模型图标（CVLM和MICV），并将它们添加到ICONS对象中。合并请求增加了对模型卡片的处理逻辑，调整了导入顺序，并在点击模型卡片时根据是否存在\`apiDoc\`属性决定是跳转到详情页还是打开新窗口。新增了两个模型卡片“通义千问-Max”和“通义千问-Plus”，包括其详细描述、标签、类型、更新时间和API文档链接。`
+  const summarizedComment = `
+  # 整体摘要：
+  ${summarizedMr}
+
+  # 文件变更：
+  | 文件路径 | 变更摘要 |
+  | --- | --- |
+  | src/components/ModelCard/index.tsx | 新增模型卡片“通义千问-Max”和“通义千问-Plus”，包括其详细描述、标签、类型、更新时间和API文档链接。 |
+  | src/components/ModelCard/index.less | 新增模型卡片“通义千问-Max”和“通义千问-Plus”的样式。 |
+  | src/components/ModelCard/icons.ts | 增加了两个新的模型图标（CVLM和MICV），并将它们添加到ICONS对象中。 |
+  `
+
+  const loadingComment = "小煎蛋正在处理您的合并请求，请稍等片刻。"
+
+  const project_id = 139032
+  const merge_request_iid = 488
+
+  const { id } = await service.gitlab.note.create2Mr(
+    project_id,
+    merge_request_iid,
+    loadingComment
+  )
+
+  await sleep(5000)
+
+  await service.gitlab.note.modify2Mr(
+    project_id,
+    merge_request_iid,
+    id,
+    summarizedComment
+  )
+}, 10000)
--- a/types/gitlab.ts
+++ b/types/gitlab.ts
@ -347,4 +347,42 @@ export namespace Gitlab {
     */
    web_url: string
  }
+
+  /* 合并请求事件 */
+  export interface MergeRequestEvent {
+    /**
+     * 事件对象类型
+     */
+    object_kind: "merge_request"
+    /**
+     * 事件类型
+     */
+    event_type: "merge_request"
+    /**
+     * 项目信息
+     */
+    project: {
+      /**
+       * 项目ID
+       */
+      id: number
+    }
+    /**
+     * 合并请求的属性
+     */
+    object_attributes: {
+      /**
+       * 合并请求的内部ID
+       */
+      iid: number
+      /**
+       * 合并请求的状态
+       */
+      state: "opened" | "closed" | "reopened" | "merged"
+      /**
+       * 合并请求的变更信息
+       */
+      changes?: any
+    }
+  }
 }
--- a/utils/chatTools.ts
+++ b/utils/chatTools.ts
@ -0,0 +1,32 @@
+import { ChatOpenAI } from "@langchain/openai"
+
+/**
+ * 获取Deepseek模型
+ * @param {number} temperature - 温度参数，用于控制生成文本的随机性。
+ * @returns {Promise<ChatOpenAI>} 返回一个包含Deepseek模型实例的Promise。
+ */
+const getDeepseekModel = async (temperature: number) => {
+  const model = "deepseek-coder"
+  const apiKey = "sk-21a2ce1c2ee94bc2933798eac1bbcadc"
+  const baseURL = "https://api.deepseek.com"
+  return new ChatOpenAI({ apiKey, temperature, model }, { baseURL })
+}
+
+/**
+ * 获取GPT-4o模型
+ * @param {number} temperature - 温度参数，用于控制生成文本的随机性。
+ * @returns {Promise<ChatOpenAI>} 返回一个包含GPT-4o模型实例的Promise。
+ */
+const getGpt4oModel = async (temperature: number) => {
+  const model = "deepseek-coder"
+  const apiKey = "sk-EhbBTR0QjhH22iLr9aCb04D2B0F44f88A07c2924Eb54CfA4"
+  const baseURL = "https://api.gpt.ge/v1"
+  return new ChatOpenAI({ apiKey, temperature, model }, { baseURL })
+}
+
+const chatTools = {
+  getDeepseekModel,
+  getGpt4oModel,
+}
+
+export default chatTools
--- a/utils/pathTools.ts
+++ b/utils/pathTools.ts
@ -30,3 +30,22 @@ export const makeCheckPathTool = (url: string, prefix?: string) => {
    fullCheck: (path: string) => pathname === path,
  }
 }
+
+/**
+ * 裁剪路径字符串，如果路径长度超过20个字符，则只保留最后两级目录。
+ *
+ * @param {string} path - 要处理的路径字符串。
+ * @returns {string} - 裁剪后的路径字符串，如果长度不超过20个字符则返回原路径。
+ */
+export const shortenPath = (path: string): string => {
+  if (path.length <= 20) {
+    return path
+  }
+
+  const parts = path.split("/")
+  if (parts.length <= 2) {
+    return path
+  }
+
+  return `.../${parts[parts.length - 2]}/${parts[parts.length - 1]}`
+}
--- a/utils/tokenTools.ts
+++ b/utils/tokenTools.ts
@ -0,0 +1,19 @@
+import { get_encoding as getEncoding } from "@dqbd/tiktoken"
+
+const tokenizer = getEncoding("cl100k_base")
+
+const encode = (input: string): Uint32Array => {
+  return tokenizer.encode(input)
+}
+
+const getTokenCount = (input: string): number => {
+  const cleanedInput = input.replace(/<\|endoftext\|>/g, "")
+  return encode(cleanedInput).length
+}
+
+const tokenTools = {
+  getTokenCount,
+  encode,
+}
+
+export default tokenTools