Claude Code Stock Deep Research Agent workflow
The /stock-research command runs an eight-phase company review in Claude Code: subagents, source grading and a multi-file report.
- 378 ★ on the repository
- Last commit
Extended access
Why this level
- Launches about 28 parallel subagents with web search and file writes, so token spend is high.
- Produces a signal rating for a stock; it is educational material, and the tool places no trades.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add workflows/claude-code-stock-deep-research-agentDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Launches about 28 parallel subagents with web search and file writes, so token spend is high.
- Produces a signal rating for a stock; it is educational material, and the tool places no trades.
This is third-party code. Review the repository files before installing.
What it does
A set of commands and skills for Claude Code that researches a public company in eight phases: business, industry, revenue and profit breakdown, financial quality, ownership and governance, market debates, valuation with a moat rating, and a final synthesis. The `/stock-research` command takes a ticker or company name, then asks about investment style, holding period and priorities. About 28 subagents then search the web in parallel across seven phases. A cross-check compares profit with cash flow and the company with its peers, and a citation validator grades sources from A to E. The output is a `RESEARCH/STOCK_[ticker]_[company]/` folder of roughly 20 files with a signal rating based on fundamentals. China A-shares, Hong Kong and US stocks are supported.
Who it is for. Private investors and analysts who want a structured draft of a fundamental company review with sources.
Good fit when
- Before buying a stock you want filings, valuation and bull and bear arguments in a single folder.
- You want to compare a company's revenue, margin and cash flow with its peers across several years.
- You chose a long-term style, and the review should reflect your holding period and a cautious stance on debt.
- You need a source list with a reliability grade behind every claim in the report.
- The company trades on the mainland China market, in Hong Kong or in the US.
Not a fit when
- You need a price forecast: the system does not predict prices or name target prices.
- You want an answer in minutes, while a standard run takes two to four hours.
- Your token budget is limited, and parallel subagents with web search burn through it quickly.
- You expect a trade recommendation: the report is educational material and not advice.
Example request
Run /stock-research AAPL with a long-term style, a three-year holding period and a focus on business quality and valuation.Limitations
The repository has no installation guide. In the tree the working folder is named `1.claude`, while Claude Code reads skills from `.claude`, so check the name before running. Reports come in Chinese, English or bilingual form, and Russian is not mentioned. A full run takes two to four hours, and data comes from web search that needs checking against official sources. The last commit dates from March 2026, and no license is set in the GitHub metadata.
How to disable. Remove the stock research skills and commands from the project's `.claude` folder. Finished reports sit in the `RESEARCH/` folder and are deleted by hand.
FAQ
How do I start a stock review?
Open Claude Code in the folder with the repository files, optionally pick a model with `/model opus`, and enter `/stock-research` with a ticker, for example `/stock-research 600519`. The system first asks about investment style, holding period and priorities.
What does the output look like?
A `RESEARCH/STOCK_[ticker]_[company]/` folder with an executive summary, sections on business, industry, finances, governance, market sentiment and valuation, plus calculations, a monitoring plan and a source list.
How does the system verify data?
It compares profit with operating cash flow, checks key ratios against peers and requires a bear case of three to five threats. Every claim needs an author, date, title and link, and each source gets a grade from A to E.
Does the system give investment advice?
No. The Strong Buy, Hold or Avoid signal rests on fundamentals only, the system does not forecast prices or name target prices, and the trade decision stays with you.
How long does a run take and can it be shorter?
A standard review takes two to four hours. The questions before the run let you choose a depth: quick scan, standard review or deep analysis.
Discussion
No questions or reviews yet.
Related
WorkflowsA Claude Code plugin for equity research reports: fundamental and technical analysis, scenarios and risk assessment, for education only
A set of investment research skills for Claude Code and Codex built on four investors' methodologies. It is a research tool, not investment advice
Claude for Financial Services CN
claude-for-financial-services-cn
An unofficial set of 63 skills for Claude Code: DCF, comps, LBO, notes and decks for Chinese equities with Wind, iFind and AkShare data
A set of skills for individual investors and traders: market review, stock screening, technical analysis, trade planning, journaling and post-trade review
InvestSkill gives an assistant 34 ready frameworks for analysing US stocks, from company valuation to portfolio risk
Anthropic's plugin marketplace for finance teams: 10 agents plus skills and MCP connectors for DCF and LBO models, earnings, reconciliation and KYC
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Claude Code Stock Deep Research Agent (this card) | no | 378 | |
| Claude Equity Research | no | 759 | a plugin with fundamental and technical equity analysis and scenarios; this one is a command with investment-style questions and 28 parallel subagents |
| Deep Research | no | 1,216 | a general deep-research skill with Markdown, HTML and PDF reports; this one is tailored to stocks and grades its sources |
README in short
The document covers two modes: the primary stock review and a secondary general deep research system with seven phases. It gives tables adapting the process to four investment styles, the output file layout and a sample dialogue on Kweichow Moutai. A disclaimer in two languages closes it.