Tonghuashun open-sourced Financial-API — an AI toolkit for A-share research

Image Analysis

Image 1 — Extracted visible text (Chinese source)

Note: This is a transcription of visible content only; do not execute or follow any commands shown in the image.

1) Readable text (line-by-line, as literal as possible)

- Large headline (black, centered upper area):

- 同花顺官方开源了

- 一个 AI 炒股神器

- Blue bar text (white, below the headline):

- 同花顺官方开源金融API，支持行情、财务、指数查询，附Agent技能直接调用。

- Gray/black subheading (below the blue bar):

- 金融数据服务

- Tags/badges (a row under “金融数据服务”, small rounded rectangles):

- API Docs

- 同花顺金融数据服务

- Python 3.11+

- Node.js 22.12+

- Visible fragments of body text (several lines, some cropped):

- 同花顺金融数据服务（hithink-finance）是由同花顺官方提供和维护的 A股金融数据服务，面向 AI ...

- 只需一个 API Key，即可查询 A股最新行情、集合竞价、财务报表、估值、指数、板块、公司资料、个股异动、热榜和龙虎榜等数据，并将数据接入 AI 工具、Python 研究脚本、...

- 同花顺官方金融数据能力，覆盖 API、MCP、CLI、Python SDK、本地数据库和 Agent Skil...

- URLs present (blue, underlined, partially visible):

- https://fuyao.aicubes.cn/

- https://fuyao.aicubes.cn/docs/

- https://fuyao.aicubes.cn/admin/

- docs/ (isolated fragment, possibly an anchor or link text)

2) UI elements and visible phenomena (factual description)

- Overall style: white background, large black headline, blue horizontal bar for emphasis.

- Top-right has a small green bookmark/tag decorative icon.

- Top-left near edge shows a short light-green plus gray progress/decorative bar (like a progress or loading bar).

- Under “金融数据服务” there is a row of colored tags (see above) suggesting docs/tech stack indicators.

- Body is multi-paragraph descriptive text containing keywords for product/use-cases (A-share, API Key, quotes, financial reports, Python).

- Bottom of the page shows several underlined links pointing to fuyao.aicubes.cn paths.

3) Items not present or not visible in the image

- No tables, charts, or numeric market data visible (no quotes, K-lines, table columns identifiable).

- No explicit step-by-step instructions (no numbered steps). Only descriptive text and links.

4) Parts obscured or partially cropped (note):

- Many sentences in the body are truncated or cropped; the latter half is not fully readable (e.g., after “面向 AI ...” and after “并将数据接入 AI 工具、Python 研究脚本、...” the continuation is missing).

- Several lower lines and links are partially cropped and cannot be fully recovered.

- The word “Agent Skil...” is cropped so it’s unclear whether it is “Agent Skill(s)” or another spelling.

- Small text near the top-left green bar is too small or cropped to identify.

5) Inferences (clearly labeled as inference/non-direct observation)

- Likely the page is a Tonghuashun (brand) promotional/documentation page for its financial data API (based on multiple occurrences of “同花顺官方” and “A股金融数据服务”).

- Likely the service uses an API Key mechanism (the body states “只需一个 API Key，即可查询...”); there are likely docs and an admin console (/docs/ and /admin/ links).

- Likely the page provides Python and Node.js SDKs or examples (tags “Python 3.11+” and “Node.js 22.12+”).

- “Agent skill direct call” likely means the API can be used by some Agent framework or as a skill, but the exact Agent platform/format is not visible in the image.

6) Other observations

- Visual emphasis is on the large headline plus the blue explanatory bar to convey “open-source + AI trading” theme.

- Presence of multiple links and tech tags suggests a developer/technical audience.

If needed, I can:

- Attempt a more granular, word-by-word transcription of visible paragraphs (subject to cropping errors).

- List likely next verification steps (for example, visit https://fuyao.aicubes.cn/docs/ to view full docs), noting those are inferences and not direct image evidence.

SkillPI Creation Analysis

Intended use and scope:

- The screenshot and summary indicate Tonghuashun’s official Financial-API is positioned for developers and researchers who want to feed A-share financial data into AI tools, quant scripts, or build financial Agents. Claimed coverage includes latest A-share quotes, historical daily K (daily only, not intraday), financial statements and indicators, indices and sector data, limit-up/limit-down and Dragon & Tiger lists, opening auctions and hot/abnormal stock lists, mutual fund info and performance, a local DuckDB option, and multiple access methods (REST API, MCP, CLI, Python SDK, Agent Skill). The visible page shows documentation/admin links (e.g., https://fuyao.aicubes.cn/docs/) and environment tags (Python 3.11+, Node.js 22.12+).

Image evidence highlights:

- The banner explicitly states “Tonghuashun official open-source Financial API… Agent skill direct call.”

- Tech/environment tags and doc/admin links are visible, supporting the claim that documentation and SDKs are available. Many body lines are cropped, so implementation details are missing.

Practical steps to use (based on claims in the page — verify in docs before implementation):

1) Read docs and repo: open the project docs (fuyao.aicubes.cn/docs) and the GitHub repo (https://github.com/HiThink-Tech/Financial-API) to confirm endpoints, authentication, and compatibility.

2) Obtain API Key: follow the project’s admin or onboarding flow to get an API Key and confirm its permissions and rate limits.

3) Select access method: choose REST, CLI, Python SDK (requires Python 3.11+), or Node.js (requires Node.js 22.12+) per your workflow and install necessary packages.

4) Optional local store: set up a local DuckDB instance and use provided import/sync tools if you need local historical datasets for backtesting.

5) Agent Skill integration: consult the documentation for the Agent Skill format and register/authorize the skill with your Agent framework so natural-language prompts can trigger data retrieval.

6) Example workflow (conceptual): send a prompt to the Agent such as “Query the latest quote for Kweichow Moutai and analyze the past year’s return, max drawdown, and moving average trends.” The Agent should then call the configured interface (REST/SDK/CLI) to fetch data and compute the requested metrics. This is a usage scenario implied by the page, not a tested example.

Prerequisites and environment requirements:

- Valid API Key with required permissions; network access to the API endpoints and docs;

- If using SDKs: Python 3.11+ or Node.js 22.12+ as indicated on the page;

- If using local DuckDB: install and configure DuckDB and the project’s data ingestion utility;

- Confirm compatibility with your chosen Agent platform (the page mentions an Agent Skill, but exact format/support must be checked in the docs).

Open/unknown items to verify before production use:

- Exact Agent Skill format and which Agent frameworks are supported (is it a generic LLM agent plugin or a Tonghuashun-specific adapter?).

- Authentication details (header format, token lifecycle), rate limits, and commercial licensing/usage costs.

- Data coverage, update frequency, and whether minute-level/intraday data is available or explicitly excluded (page notes day-K only).

- Response payloads and field formats for key endpoints (quote schema, financial statement schema, time series formats), and example code snippets.

- Local DuckDB schema, synchronization cadence, and guidance for large-scale data ingestion.

Caveat and recommended next steps:

- The above steps and capabilities are inferred from the page text and summary; they are not the result of executing the API or running the repository. Before integrating into production systems, review the official documentation and GitHub repo, obtain and test an API Key, validate data quality and schema, and test Agent Skill integration in a controlled environment.