knowme-careerforge-skill 0.0.7 → 0.0.8

package/scripts/evidence/extract-evidence.py

@@ -1,395 +1,21 @@
 #!/usr/bin/env python3
-"""
-KnowMe CareerForge — Codebase & Git Evidence Miner (repo-to-resume-tailor engine)
-深度解析本地代码工程、依赖清单、系统架构与 Git 提交历史，提取 L1~L3 结构化证据链并输出 Candidate Master Profile。
-"""
+"""Retired compatibility entry: repository contents are not candidate evidence."""
+import json
+import sys
 
-import sys, os, json, argparse, re, subprocess
-from pathlib import Path
-from typing import Dict, List, Any, Optional
-
-def get_project_root() -> Path:
-    curr = Path(__file__).resolve().parent
-    for _ in range(5):
-        if (curr / "SKILL.md").exists() or (curr / "package.json").exists():
-            return curr
-        curr = curr.parent
-    return Path.cwd()
-
-TECH_SIGNATURES = {
-    # AI / LLM & Agents
-    "LangChain": ["langchain", "langchain-core", "@langchain/core"],
-    "LangGraph": ["langgraph", "@langchain/langgraph"],
-    "LlamaIndex": ["llama-index", "llamaindex"],
-    "Qdrant": ["qdrant-client", "@qdrant/js-client-rest"],
-    "Milvus": ["pymilvus", "@zilliz/milvus2-sdk-node"],
-    "FastAPI": ["fastapi", "uvicorn"],
-    "PyTorch": ["torch", "torchvision"],
-    "OpenAI SDK": ["openai", "@openai/openai"],
-    # Web & Fullstack
-    "TypeScript": ["typescript", "ts-node", "@types/node"],
-    "React": ["react", "react-dom", "next"],
-    "Next.js": ["next"],
-    "Vue.js": ["vue", "nuxt"],
-    "Node.js": ["express", "koa", "nestjs", "@nestjs/core", "fastify"],
-    "Tailwind CSS": ["tailwindcss", "@tailwindcss/postcss"],
-    "Playwright": ["playwright", "@playwright/test"],
-    "GraphQL": ["graphql", "apollo-server", "@apollo/client"],
-    # Backend & Systems
-    "Go / Golang": ["github.com/gin-gonic/gin", "google.golang.org/grpc"],
-    "Rust": ["tokio", "serde", "actix-web", "axum"],
-    "Spring Boot": ["spring-boot", "org.springframework.boot"],
-    # Database & Cache
-    "PostgreSQL": ["pg", "psycopg2", "asyncpg", "typeorm", "prisma"],
-    "MySQL": ["mysql", "mysql2", "pymysql"],
-    "Redis": ["redis", "ioredis", "aioredis"],
-    "Elasticsearch": ["elasticsearch", "@elastic/elasticsearch"],
-    # DevOps & Infrastructure
-    "Docker": ["Dockerfile", "docker-compose.yml", "docker-compose.yaml"],
-    "Kubernetes": ["k8s", "helm", "Deployment", "StatefulSet"],
-    "GitHub Actions": [".github/workflows"]
-}
-
-def run_cmd(cmd: List[str], cwd: Optional[Path] = None) -> str:
-    try:
-        res = subprocess.run(cmd, cwd=str(cwd) if cwd else None, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True, timeout=10)
-        return res.stdout.strip() if res.returncode == 0 else ""
-    except Exception:
-        return ""
-
-def scan_repo(repo_path: Path) -> Dict[str, Any]:
-    repo_path = repo_path.resolve()
-    facts: Dict[str, Any] = {
-        "repo_name": repo_path.name,
-        "repo_path": str(repo_path),
-        "detected_languages": {},
-        "detected_tech_stacks": set(),
-        "architectural_signals": [],
-        "author_info": {},
-        "git_stats": {},
-        "key_dependencies": [],
-        "evidence_items": []
-    }
-
-    if not repo_path.exists() or not repo_path.is_dir():
-        return facts
-
-    # 1. Git 信息提取 (Author, commits, tags)
-    git_dir = repo_path / ".git"
-    if git_dir.exists():
-        user_name = run_cmd(["git", "config", "user.name"], cwd=repo_path)
-        user_email = run_cmd(["git", "config", "user.email"], cwd=repo_path)
-        commit_count = run_cmd(["git", "rev-list", "--count", "HEAD"], cwd=repo_path)
-        remote_url = run_cmd(["git", "config", "--get", "remote.origin.url"], cwd=repo_path)
-        recent_commits = run_cmd(["git", "log", "-n", "8", "--pretty=format:%s"], cwd=repo_path)
-
-        facts["author_info"]["name"] = user_name or ""
-        facts["author_info"]["email"] = user_email or ""
-        facts["author_info"]["git_remote"] = remote_url or ""
-        facts["git_stats"]["total_commits"] = int(commit_count) if commit_count.isdigit() else 0
-        facts["git_stats"]["recent_commits"] = [c for c in recent_commits.splitlines() if c]
-
-    # 2. 统计文件类型分布
-    lang_map = {
-        ".ts": "TypeScript", ".tsx": "TypeScript/React", ".js": "JavaScript", ".jsx": "JavaScript/React",
-        ".py": "Python", ".go": "Go", ".rs": "Rust", ".java": "Java", ".cpp": "C++", ".c": "C",
-        ".html": "HTML", ".css": "CSS", ".vue": "Vue", ".sql": "SQL", ".sh": "Shell"
-    }
-    
-    file_counts = {}
-    ignored_dirs = {".git", "node_modules", ".venv", "venv", "__pycache__", "dist", "build", ".next", ".turbo"}
-    
-    for root, dirs, files in os.walk(repo_path):
-        dirs[:] = [d for d in dirs if d not in ignored_dirs and not d.startswith(".")]
-        for f in files:
-            p = Path(root) / f
-            ext = p.suffix.lower()
-            if ext in lang_map:
-                lang = lang_map[ext]
-                file_counts[lang] = file_counts.get(lang, 0) + 1
-
-    facts["detected_languages"] = file_counts
-
-    # 3. 依赖文件深度解析
-    dep_names = set()
-    
-    # 3.1 Node/TS
-    pkg_json = repo_path / "package.json"
-    if pkg_json.exists():
-        try:
-            pkg_data = json.loads(pkg_json.read_text(encoding="utf-8"))
-            if not facts["author_info"].get("name") and pkg_data.get("author"):
-                author_val = pkg_data.get("author")
-                facts["author_info"]["name"] = author_val if isinstance(author_val, str) else author_val.get("name", "")
-            if not facts["author_info"].get("git_remote") and pkg_data.get("repository"):
-                repo_val = pkg_data.get("repository")
-                facts["author_info"]["git_remote"] = repo_val if isinstance(repo_val, str) else repo_val.get("url", "")
-            
-            deps = {**pkg_data.get("dependencies", {}), **pkg_data.get("devDependencies", {})}
-            dep_names.update(deps.keys())
-        except Exception:
-            pass
-
-    # 3.2 Python
-    req_txt = repo_path / "requirements.txt"
-    if req_txt.exists():
-        try:
-            for line in req_txt.read_text(encoding="utf-8").splitlines():
-                line = line.strip()
-                if line and not line.startswith("#"):
-                    dep_names.add(re.split(r'[=<>~]', line)[0].strip().lower())
-        except Exception:
-            pass
-
-    pyproject = repo_path / "pyproject.toml"
-    if pyproject.exists():
-        try:
-            content = pyproject.read_text(encoding="utf-8").lower()
-            for key in ["fastapi", "torch", "langchain", "qdrant", "pytest", "pydantic"]:
-                if key in content:
-                    dep_names.add(key)
-        except Exception:
-            pass
-
-    # 3.3 Go
-    gomod = repo_path / "go.mod"
-    if gomod.exists():
-        try:
-            for line in gomod.read_text(encoding="utf-8").splitlines():
-                line = line.strip()
-                if line.startswith("require") or " " in line:
-                    parts = line.split()
-                    if len(parts) >= 2 and "/" in parts[0]:
-                        dep_names.add(parts[0])
-        except Exception:
-            pass
-
-    # 3.4 Rust
-    cargo = repo_path / "Cargo.toml"
-    if cargo.exists():
-        try:
-            for line in cargo.read_text(encoding="utf-8").splitlines():
-                line = line.strip()
-                if "=" in line and not line.startswith("["):
-                    dep_names.add(line.split("=")[0].strip().lower())
-        except Exception:
-            pass
-
-    # 4. 匹配技术栈标签
-    for tech, sigs in TECH_SIGNATURES.items():
-        for sig in sigs:
-            if sig.lower() in [d.lower() for d in dep_names] or (repo_path / sig).exists():
-                facts["detected_tech_stacks"].add(tech)
-                break
-
-    # 5. 架构与工程化信号
-    if (repo_path / "packages").exists() or (repo_path / "pnpm-workspace.yaml").exists():
-        facts["architectural_signals"].append("Monorepo 模块化架构")
-    if (repo_path / "Dockerfile").exists() or (repo_path / "docker-compose.yml").exists():
-        facts["architectural_signals"].append("Docker 容器化交付")
-    if (repo_path / ".github" / "workflows").exists():
-        facts["architectural_signals"].append("CI/CD 自动化流水线")
-    if (repo_path / "tests").exists() or (repo_path / "test").exists():
-        facts["architectural_signals"].append("自动化单元/集成测试")
-
-    facts["key_dependencies"] = sorted(list(dep_names))
-    facts["detected_tech_stacks"] = sorted(list(facts["detected_tech_stacks"]))
-
-    # 6. 生成 L1~L3 证据项
-    evidence_items = []
-    if facts["detected_tech_stacks"]:
-        evidence_items.append({
-            "claim": f"基于 {', '.join(facts['detected_tech_stacks'][:4])} 构建核心工程架构，实现高可靠端到端业务流",
-            "evidenceLevel": "L1",
-            "evidenceSource": f"代码库依赖配置文件 ({', '.join([f for f in ['package.json', 'go.mod', 'requirements.txt', 'Cargo.toml'] if (repo_path / f).exists()]) or '依赖树'})",
-            "keywords": facts["detected_tech_stacks"][:5]
-        })
-
-    if facts["architectural_signals"]:
-        evidence_items.append({
-            "claim": f"主导设计工程架构体系，落地 {', '.join(facts['architectural_signals'])}，保障交付效率与质量",
-            "evidenceLevel": "L1",
-            "evidenceSource": "工程根目录架构配置文件与流水线定义",
-            "keywords": ["Architecture", "CI/CD", "Testing", "DevOps"]
-        })
-
-    if facts["git_stats"].get("total_commits", 0) > 10:
-        evidence_items.append({
-            "claim": f"持续推进工程重构与版本迭代，累计提交 {facts['git_stats']['total_commits']} 次代码并完成多版本发布",
-            "evidenceLevel": "L2",
-            "evidenceSource": "Git 提交历史与版本变更记录",
-            "keywords": ["Refactoring", "Git", "Release Engineering"]
-        })
-
-    facts["evidence_items"] = evidence_items
-    return facts
-
-def build_master_profile(facts: Dict[str, Any], user_overrides: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
-    """Build structured candidate profile from codebase facts + optional user overrides.
-
-    Placeholder discipline:
-    - repo_mode=True : facts dict is populated from real repo scan; safe to synthesize
-      experience/education skeletons because they'll be user-reviewed before use.
-    - repo_mode=False: only user_overrides carry real data. NEVER emit fabricated placeholder
-      strings (fake company, fake school, fake dates). Use null / empty list instead.
-    """
-    overrides = user_overrides or {}
-    author = facts.get("author_info", {})
-    repo_mode: bool = bool(facts.get("detected_tech_stacks") or facts.get("evidence_items"))
-
-    # ── Basics ────────────────────────────────────────────────────────────────
-    name  = overrides.get("name")  or author.get("name")  or (None if not repo_mode else "候选人")
-    title = overrides.get("title") or (None if not repo_mode else "资深工程师 / 技术专家")
-    email = overrides.get("email") or author.get("email") or (None if not repo_mode else "candidate@example.com")
-    phone = overrides.get("phone") or (None if not repo_mode else None)
-    github = overrides.get("github") or author.get("git_remote") or None
-    location = overrides.get("location") or (None if not repo_mode else None)
-
-    tech_stacks = facts.get("detected_tech_stacks", [])
-    languages   = list(facts.get("detected_languages", {}).keys())
-
-    # ── Skills ────────────────────────────────────────────────────────────────
-    all_tech = tech_stacks + [l for l in languages if l not in tech_stacks]
-    arch_signals = facts.get("architectural_signals", [])
-    skills: list = []
-    if all_tech:
-        skills.append({"category": "核心技术栈", "items": all_tech, "highlighted": all_tech[:4]})
-    if arch_signals:
-        skills.append({"category": "工程化与架构", "items": arch_signals, "highlighted": arch_signals[:2]})
-
-    # ── Experience ────────────────────────────────────────────────────────────
-    bullets = [
-        {
-            "text": item["claim"],
-            "evidenceLevel": item["evidenceLevel"],
-            "evidenceSource": item["evidenceSource"],
-            "keywords": item.get("keywords", [])
-        }
-        for item in facts.get("evidence_items", [])
-    ]
-
-    if repo_mode and bullets:
-        # Repo mode: synthesize a single experience entry from code evidence.
-        # Company / dates are placeholders that the user MUST review.
-        experience = [
-            {
-                "company": overrides.get("company") or "[请填写：公司名称]",
-                "role": title or "[请填写：职位名称]",
-                "startDate": overrides.get("startDate") or "[请填写：开始时间]",
-                "endDate": overrides.get("endDate") or "至今",
-                "location": location or "[请填写：工作地点]",
-                "summary": "负责核心产品研发与关键架构攻坚",
-                "bullets": bullets
-            }
-        ]
-    elif overrides.get("experience"):
-        experience = overrides["experience"]
-    else:
-        experience = []
-
-    # ── Projects ──────────────────────────────────────────────────────────────
-    if repo_mode and tech_stacks:
-        projects = [
-            {
-                "name": facts.get("repo_name") or "[请填写：项目名称]",
-                "role": "核心负责人 / 主力开发者",
-                "techStack": tech_stacks[:5],
-                "repoUrl": github,
-                "startDate": "[请填写：开始时间]",
-                "endDate": "至今",
-                "bullets": [
-                    {"text": item["claim"], "evidenceLevel": item["evidenceLevel"], "evidenceSource": item["evidenceSource"]}
-                    for item in facts.get("evidence_items", [])
-                ]
-            }
-        ]
-    elif overrides.get("projects"):
-        projects = overrides["projects"]
-    else:
-        projects = []
-
-    # ── Education ─────────────────────────────────────────────────────────────
-    if overrides.get("education"):
-        education = overrides["education"]
-    elif overrides.get("school") or overrides.get("degree"):
-        education = [
-            {
-                "institution": overrides.get("school") or "[请填写：学校名称]",
-                "degree": overrides.get("degree") or "[请填写：学历与专业]",
-                "startDate": overrides.get("edu_start") or "[请填写：入学时间]",
-                "endDate": overrides.get("edu_end") or "[请填写：毕业时间]"
-            }
-        ]
-    elif repo_mode:
-        # Repo mode: emit a reminder skeleton so the user knows to fill it in.
-        education = [
-            {
-                "institution": "[请填写：学校名称]",
-                "degree": "[请填写：学历与专业]",
-                "startDate": "[请填写：入学时间]",
-                "endDate": "[请填写：毕业时间]"
-            }
-        ]
-    else:
-        education = []
-
-    # ── Assemble ──────────────────────────────────────────────────────────────
-    basics: Dict[str, Any] = {}
-    if name:    basics["name"]     = name
-    if title:   basics["title"]    = title
-    if email:   basics["email"]    = email
-    if phone:   basics["phone"]    = phone
-    if location: basics["location"] = location
-    if github:  basics["github"]   = github
-    if all_tech:
-        basics["summary"] = (
-            f"拥有坚实的技术研发与工程化落地实战经验，"
-            f"主导过多个基于 {', '.join(all_tech[:3])} 的核心系统设计与交付。"
-        )
-
-    profile: Dict[str, Any] = {"basics": basics}
-    if skills:     profile["skills"]     = skills
-    if experience: profile["experience"] = experience
-    if projects:   profile["projects"]   = projects
-    if education:  profile["education"]  = education
-    return profile
 
 def main():
-    parser = argparse.ArgumentParser(description="KnowMe CareerForge — Codebase Evidence Miner")
-    parser.add_argument("--repo", "-r", default=".", help="Target repository directory (default: current directory)")
-    parser.add_argument("--output", "-o", default="workspace/evidence-master.json", help="Output evidence JSON path")
-    parser.add_argument("--name", help="Candidate full name override")
-    parser.add_argument("--role", help="Candidate target role/title override")
-    parser.add_argument("--email", help="Candidate email override")
-    parser.add_argument("--phone", help="Candidate phone override")
-    parser.add_argument("--quiet", "-q", action="store_true", help="Quiet execution mode")
-
-    args = parser.parse_args()
-    repo_path = Path(args.repo)
-
-    facts = scan_repo(repo_path)
-    
-    overrides = {}
-    if args.name: overrides["name"] = args.name
-    if args.role: overrides["title"] = args.role
-    if args.email: overrides["email"] = args.email
-    if args.phone: overrides["phone"] = args.phone
-
-    profile = build_master_profile(facts, overrides)
-
-    out_path = Path(args.output)
-    out_path.parent.mkdir(parents=True, exist_ok=True)
-    out_path.write_text(json.dumps(profile, ensure_ascii=False, indent=2), encoding="utf-8")
-
-    if not args.quiet:
-        print("==============================================================")
-        print("  KnowMe CareerForge — Codebase Evidence Mining Report")
-        print(f"  Target Repository : {facts['repo_path']}")
-        print(f"  Detected Techs    : {', '.join(facts['detected_tech_stacks']) or 'None'}")
-        print(f"  Architecture      : {', '.join(facts['architectural_signals']) or 'Standard'}")
-        print(f"  Evidence Generated: {len(facts['evidence_items'])} L1/L2 items")
-        print(f"  Profile Written To: {out_path}")
-        print("==============================================================")
-
-if __name__ == "__main__":
-    main()
+    # Intentionally do not parse repository/output paths or read candidate inputs.
+    # Old CLI/npm/direct callers must fail before creating or overwriting files.
+    print(json.dumps({
+        'status': 'FAIL',
+        'stage': 'unsupported',
+        'errors': ['Repository-to-resume extraction is not supported. Use user-provided facts with forge --profile-json.'],
+        'warnings': [],
+        'outputs': {},
+    }))
+    return 1
+
+
+if __name__ == '__main__':
+    sys.exit(main())