"<line_number_rules>" "Each line in the code context is prefixed with its actual line number (e.g., '100. code here'). " "When citing source lines, read the line numbers directly from these prefixes. " "Do not count or calculate line numbers yourself. " "</line_number_rules>"
deffind_proto_files(repo_path: str, excluded_dirs=None) -> List[str]: """遍历 repo 目录，返回所有 .proto 文件路径""" skip = set(DEFAULT_EXCLUDED_DIRS) # 排除 vendor、node_modules 等 proto_files = [] for root, dirs, files in os.walk(repo_path): dirs[:] = [d for d in dirs if d notin skip] for f in files: if f.endswith(".proto"): proto_files.append(os.path.join(root, f)) proto_files.sort() return proto_files
“Skills are organized folders of instructions, scripts, and resources that agents can discover and load dynamically to perform better at specific tasks.”
可以把 Skills 理解为”分类后的系统提示词”，但它比传统的 System Prompt 更智能——按需加载，而不是全量加载。
维度
传统系统提示词
Agent Skills
加载方式
全量加载：每次对话都要发一遍
按需调用：只加载需要的技能
Token 消耗
高：Prompt 长度随功能增多而爆炸
低：结合 Prompt Caching 降低成本
复杂度上限
低：Prompt 太长会”注意力失焦”
高：每个技能独立，互不干扰
执行能力
仅限”说话”
可关联 Tool Use，真正执行操作
Skills 的本质：提示词工程的进化
说到底，Skills 的本质还是前几年流行的提示词工程（Prompt Engineering）。
回想一下 2023 年 ChatGPT 刚火的时候，网上到处都是”万能提示词模板”、”让 AI 效率翻倍的 prompt 技巧”。那时候大家都在研究怎么写出更好的 System Prompt，让 AI 扮演各种角色：翻译官、程序员、文案专家…
Skills 做的事情本质上没变——**还是在告诉 AI “你是谁、你能做什么、你应该怎么做”**。
defon_check_suite(request: CheckRunCompleted): logger.info(f"Received check run completed event for {request.repository.full_name}") g = get_github_client(request.installation.id) repo = g.get_repo(request.repository.full_name) ifnot get_gha_enabled(repo): logger.info(f"Skipping github action for {request.repository.full_name} because it is not enabled") returnNone pr = repo.get_pull(request.check_run.pull_requests[0].number) num_pr_commits = len(list(pr.get_commits())) if num_pr_commits > 20: logger.info(f"Skipping github action for PR with {num_pr_commits} commits") returnNone logger.info(f"Running github action for PR with {num_pr_commits} commits") logs = download_logs( request.repository.full_name, request.check_run.run_id, request.installation.id ) ifnot logs: returnNone logs = clean_logs(logs) extractor = GHAExtractor() logger.info(f"Extracting logs from {request.repository.full_name}, logs: {logs}") problematic_logs = extractor.gha_extract(logs) if problematic_logs.count(" ") > 15: problematic_logs += "
========================================
There are a lot of errors. This is likely a larger issue with the PR and not a small linting/type-checking issue." comments = list(pr.get_issue_comments()) iflen(comments) >= 2and problematic_logs == comments[-1].body and comments[-2].body == comments[-1].body: comment = pr.as_issue().create_comment(log_message.format(error_logs=problematic_logs) + "
I'm getting the same errors 3 times in a row, so I will stop working on fixing this PR.") logger.warning("Skipping logs because it is duplicated") raise Exception("Duplicate error logs") print(problematic_logs) comment = pr.as_issue().create_comment(log_message.format(error_logs=problematic_logs)) on_comment( repo_full_name=request.repository.full_name, repo_description=request.repository.description, comment=problematic_logs, pr_path=None, pr_line_position=None, username=request.sender.login, installation_id=request.installation.id, pr_number=request.check_run.pull_requests[0].number, comment_id=comment.id, repo=repo, ) return {"success": True}
defon_check_suite(request: CheckRunCompleted): logger.info(f"Received check run completed event for {request.repository.full_name}") g = get_github_client(request.installation.id) repo = g.get_repo(request.repository.full_name) ifnot get_gha_enabled(repo): logger.info(f"Skipping github action for {request.repository.full_name} because it is not enabled") returnNone pr = repo.get_pull(request.check_run.pull_requests[0].number) num_pr_commits = len(list(pr.get_commits())) if num_pr_commits > 20: logger.info(f"Skipping github action for PR with {num_pr_commits} commits") returnNone logger.info(f"Running github action for PR with {num_pr_commits} commits") logs = download_logs( request.repository.full_name, request.check_run.run_id, request.installation.id ) ifnot logs: returnNone logs = clean_logs(logs) extractor = GHAExtractor() logger.info(f"Extracting logs from {request.repository.full_name}, logs: {logs}") problematic_logs = extractor.gha_extract(logs) if problematic_logs.count("\n") > 15: problematic_logs += "\n\nThere are a lot of errors. This is likely a larger issue with the PR and not a small linting/type-checking issue." comments = list(pr.get_issue_comments())
==========
iflen(comments) >= 2and problematic_logs == comments[-1].body and comments[-2].body == comments[-1].body: comment = pr.as_issue().create_comment(log_message.format(error_logs=problematic_logs) + "\n\nI'm getting the same errors 3 times in a row, so I will stop working on fixing this PR.") logger.warning("Skipping logs because it is duplicated") raise Exception("Duplicate error logs") print(problematic_logs) comment = pr.as_issue().create_comment(log_message.format(error_logs=problematic_logs)) on_comment( repo_full_name=request.repository.full_name, repo_description=request.repository.description, comment=problematic_logs, pr_path=None, pr_line_position=None, username=request.sender.login, installation_id=request.installation.id, pr_number=request.check_run.pull_requests[0].number, comment_id=comment.id, repo=repo, )
language_names = ["python", "java", "cpp", "go", "rust", "ruby", "php"] # and more
# Installing the parsers languages = {} for language in LANGUAGE_NAMES: subprocess.run(f"git clone https://github.com/tree-sitter/tree-sitter-{language} cache/tree-sitter-{language}", shell=True) for language in LANGUAGE_NAMES: Language.build_library(f'cache/build/{language}.so', [f"cache/tree-sitter-{language}"]) self.languages = {language: Language(f"cache/build/{language}.so", language) for language in LANGUAGE_NAMES}
defchunk(text: str, MAX_CHARS: int = 1500) -> list[str]: # Determining the language for language_name in language_names: language = languages[language_name] parser = Parser() parser.set_language(language) tree = parser.parse(bytes(text, "utf-8")) ifnot tree.root_node.children or tree.root_node.children[0].type != "ERROR": file_language = language break logger.warning(f"Not language {language_name}")
# Smart chunker if file_language: return chunk_node(tree.root_node, text, max_chunk_size)
# Choose encoding based on embedder type if embedder_type == 'ollama': # Ollama typically uses cl100k_base encoding encoding = tiktoken.get_encoding("cl100k_base") elif embedder_type == 'google': # Google uses similar tokenization to GPT models for rough estimation encoding = tiktoken.get_encoding("cl100k_base") else: # OpenAI or default # Use OpenAI embedding model encoding encoding = tiktoken.encoding_for_model("text-embedding-3-small")
mc alias set myminio http://172.18.0.10:9000 miniouser miniopassword; mc admin user svcacct add --access-key AAAAAAAAAAAAAAAAAAAA --secret-key BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB myminio miniouser
// printSum implementation using the proposed try mechanism. funcprintSum(a, b string)error { // use a defer statement to augment errors before returning x := try(strconv.Atoi(a)) y := try(strconv.Atoi(b)) fmt.Println("result:", x + y) returnnil }
当时我们没有更好的解决方案，几年来都没有为错误处理追求语法变更。不过，社区中的许多人受到了启发，我们收到了源源不断的错误处理提案，其中许多非常相似，有些有趣，有些难以理解，有些不可行。为了跟踪不断扩大的提案，一年后，Ian Lance Taylor 创建了一个总体问题，总结了改进错误处理的提议变更的当前状态。创建了一个Go Wiki来收集相关的反馈、讨论和文章。
关于错误处理冗长性的抱怨持续存在（参见2024年上半年Go开发者调查结果），因此，在Go团队内部提案经过一系列日益完善之后，Ian Lance Taylor 在2024年发布了“使用?减少错误处理样板代码”。这次的想法是借鉴Rust中实现的构造，特别是?操作符。希望通过依靠使用既定符号的现有机制，并考虑我们多年来学到的东西，我们应该能够最终取得一些进展。在一小批用户调研中，向开发者展示使用 ? 的 Go 代码时，绝大多数参与者正确猜出了代码的含义，这进一步说服我们再试一次。为了能够看到变化的影响，Ian 编写了一个工具，将普通 Go 代码转换为使用提议的新语法的代码，我们还在编译器中对该功能进行了原型设计。
1 2 3 4 5 6 7
// printSum implementation using the proposed "?" statements. funcprintSum(a, b string)error { x := strconv.Atoi(a) ? y := strconv.Atoi(b) ? fmt.Println("result:", x + y) returnnil }
funcprintSum(a, b string)error { x, err1 := strconv.Atoi(a) y, err2 := strconv.Atoi(b) if err := cmp.Or(err1, err2); err != nil { return err } fmt.Println("result:", x+y) returnnil }
编写、阅读和调试代码都是完全不同的工作。编写重复的错误检查可能很乏味，但今天的 IDE 提供了强大的、甚至是 LLM 辅助的代码补全。编写基本的错误检查对这些工具来说很简单。在阅读代码时冗长性最明显，但工具在这里也可能有所帮助；例如，有 Go 语言设置的 IDE 可以提供一个切换开关来隐藏错误处理代码。
){kind=link}