Pans Tech Profile

分析目标公司的技术栈和算力需求，为 AI 算力销售提供客户画像。 自动采集公司官网、招聘页面、GitHub、新闻等数据源，输出技术栈分析、 GPU 需求量级估算、推理/训练比例及销售切入建议。 触发词：技术栈分析、公司技术画像、tech profile、算力需求评估、客户技术调研

Security Scan

The skill mostly implements website scraping and keyword-based tech detection, but its documentation over-promises features (GitHub, BuiltWith/news integration and automatic dependency installation) that the included script does not implement — this mismatch warrants caution.

Purpose & Capability

Skill description/README claims automated collection from multiple sources (company website, recruiting pages, GitHub org, BuiltWith/Wappalyzer, public news) and automatic dependency installation. The shipped script only fetches the target domain (and a few subpaths when --deep is used) and performs keyword matching on the fetched HTML. It does not call GitHub, BuiltWith, news APIs, or implement auto-installation. While the core purpose (profile a company from web pages) is implemented, several advertised capabilities are missing — an incoherence between declared purpose and actual capability.

Instruction Scope

SKILL.md tells the agent/user to run the Python script with company or domain parameters; the instructions do not ask for unrelated files, credentials, or system state. However the SKILL.md states "首次运行自动安装依赖" (first run auto-installs dependencies) which the script does not perform (it only prints a pip install suggestion).

Install Mechanism

There is no install spec (instruction-only). The README claims automatic dependency installation but the script merely checks for requests/bs4 and prints a pip install command if missing — it does not download or install anything. This is inconsistent but not directly risky.

Credentials

The skill requests no environment variables, no credentials, and no config paths. The script performs unauthenticated HTTP GETs only. The lack of requested secrets is proportionate to the implemented functionality.

Persistence & Privilege

The skill does not request persistent presence (always:false) and does not modify other skills or system-wide settings. It only runs as an on-demand script.

What to consider before installing

This skill will make HTTP requests to the domain you provide (or to companyname.com if you only give a name) and extract technology keywords from the returned HTML. Before installing/using: 1) Note the README over-promises features — the script does NOT query GitHub, BuiltWith, or news sources, and it does NOT auto-install dependencies; update the script yourself if you need those sources. 2) Be aware it will perform web requests to inferred domains (which may target incorrect or unintended sites); always pass explicit domains rather than just company names. 3) Because it scrapes public pages, confirm you are allowed to crawl the target site and avoid internal/private domains or credentials. 4) If you need the missing capabilities (GitHub/news/BuiltWith), review and extend the code or obtain a version that implements them. 5) If you are uncomfortable with automated scraping, run the script in a controlled environment or inspect/modify the code first.

概述

分析目标公司的技术栈和算力需求，为 AI 算力销售提供客户画像。

触发词

技术栈分析、公司技术画像、tech profile、算力需求评估、客户技术调研

使用方法

基本用法

python3 scripts/profile.py --company "OpenAI"
python3 scripts/profile.py --domain openai.com

指定输出格式

python3 scripts/profile.py --company "字节跳动" --format json
python3 scripts/profile.py --company "Anthropic" --format markdown

深度分析

python3 scripts/profile.py --company "Midjourney" --deep

参数

参数	说明	必填	默认值
--company	公司名称	二选一	-
--domain	公司域名	二选一	-
--format	输出格式：text/json/markdown	否	text
--deep	深度分析（更多数据源）	否	false
--output	输出到文件	否	stdout

输出内容

1. 公司基本信息 — 名称、域名、行业
2. 技术栈分析 — 编程语言、框架、云服务、ML/DL 工具链
3. 算力需求评估 — GPU 需求量级、推理/训练比例、预估月成本
4. 信号来源 — 官网、招聘信息、GitHub、新闻报道
5. 销售建议 — 优先级评级、切入角度

数据来源

• 公司官网 meta / tech stack 检测
• 招聘页面（技术岗位关键词）
• GitHub organization（语言分布、star 项目）
• BuiltWith / Wappalyzer 线索
• 公开新闻与融资信息

依赖

• Python 3.10+
• requests
• beautifulsoup4

首次运行自动安装依赖。