Baichuan Ai

整理并摘要百川大模型公开页面的模型能力、定价、文档目录与示例等信息，提供轻量级内容汇总服务。

Security Scan

This is an instruction-only skill that describes scraping and summarizing public Baichuan.ai pages, and it requests no installs, credentials, or privileged access — the declared behavior matches the requirements.

Purpose & Capability

The name/description (summarize public Baichuan model pages) aligns with what the skill asks for: no credentials, no installs, and only public web content extraction. Nothing requested appears unrelated to the stated purpose.

Instruction Scope

SKILL.md limits actions to publicly available pages, extraction of model/version/pricing/docs, and explicitly excludes account/API operations. It also notes waiting for dynamic loads (i.e., JS-rendered pages) and rate control. This is appropriately scoped, although dynamic rendering may require a headless browser or JS-capable fetcher in the agent environment (an implementation detail to verify).

Install Mechanism

No install spec or code files are present (instruction-only), so nothing is written to disk or fetched during install; lowest-risk posture for installation.

Credentials

No environment variables, credentials, or config paths are requested. The skill does not ask for unrelated secrets or elevated access — proportional to its purpose.

Persistence & Privilege

always is false and autonomous invocation is not disabled (the platform default). The skill does not request permanent presence or modify other skills/config; no excessive privileges are declared.

Assessment

This skill appears coherent and limited to summarizing public pages. Before installing, confirm you are comfortable with the agent making outbound web requests to baichuan.ai (and that this complies with your policies). Note: scraping JS-heavy pages may require a renderer or headless browser in the agent environment — if your agent provides that, be aware it increases the runtime surface. Verify the agent has no unrelated credentials available in its environment (to avoid accidental leakage), respect robots.txt and Baichuan's terms, and configure sensible rate limits to avoid being blocked. If you need stronger guarantees, require explicit logging/auditing of what pages are fetched and extracted.

百川智能

用途与边界

• 面向公开产品页与文档中心的检索与内容摘要
• 不提供账号操作、接口调用或权限绕过能力
• 仅处理公开页面的轻量信息整理

关键入口

• 主页：https://www.baichuan.ai/
• 文档与开放平台：站点入口
• 定价与公告：站点入口

常见任务

• 模型能力与版本说明摘要（上下文长度、功能特性）
• 定价与调用配额说明提取（公开信息）
• 文档目录与示例链接汇总

数据字段

• 模型名称、版本、上下文长度、功能特性、链接
• 定价页链接、免费额度说明、地域与服务状态
• 文档标题、目录链接、示例入口

自动化要点

• 页面动态加载，需等待完成后解析
• 不进行API调用或账号相关操作
• 频率控制，尊重平台访问限制

示例流程

• 能力摘要：访问模型页 → 抽取版本与特性 → 输出摘要
• 定价整理：访问定价入口 → 提取费用与额度 → 输出清单
• 文档汇总：进入文档中心 → 抽取目录与示例 → 输出链接集合

合规提示

• 遵守平台规则与账号安全要求
• 不存储或输出敏感信息