Stock Idea Screener
Find new long and short ideas through quantitative screens, thematic research, and pattern recognition.
The skill describes research and screening only, without requesting trading, account access, or other high-privilege actions, and the repository README requires professional review and disclaims investment advice; however, it does not specify data sources, sensitive-data handling, user confirmation, external effects, rollback, or result attribution, so points are deducted.
The workflow and screening metrics are broadly internally consistent, and the output structure is explicit; however, there are no specified tools or dependencies, missing-data and abnormal-input handling, failure feedback, reproduction method, or skill-specific tests, so the score remains moderate-low under static calibration.
The candidate identifies scenarios, trigger phrases, input parameters, screen types, and output format; however, non-fit boundaries, data-freshness requirements, regional differences, Chinese-language operation, and mainland-China reachability are not defined, warranting deductions.
The documentation is readable and layered, with workflow, metric examples, and limitations; the repository supplies an Apache-2.0 license, official-organization provenance, general CI validation, and contribution/update guidance, but the skill lacks independent versioning, changelog, maintenance ownership, installation dependencies, and troubleshooting guidance.
The workflow can produce a candidate shortlist, comparison table, and follow-up research directions, giving it clear core utility; however, no actual outputs, data integration, or quality validation are shown, and results still require fundamental research and human review, so only limited static effectiveness is supported.
The skill provides auditable rules and a presentation template, while the repository includes general plugin-validation and secret-scan CI; however, there are no skill-specific test cases, real execution evidence, source citations, or corroborating validation, so evidence coverage is limited.
- The screening metrics depend on unspecified data providers, definitions, and timestamps; verify source provenance, calculation conventions, and freshness before use.
- Outputs are research candidates rather than investment conclusions; complete fundamental, valuation, catalyst, risk, and compliance review with qualified professional sign-off.
- Chinese-language operation and mainland-China reachability for data services are unspecified; validate environmental compatibility before regional deployment.
What it does & when to use it
Idea Generation is a skill in the repository’s equity-research vertical plugin for systematic stock screening and investment-idea sourcing. It first defines search parameters such as direction, market capitalization, sector, style, geography, and theme, then organizes screens for value, growth, quality, short, or special-situation candidates. For thematic work, it develops the thesis, maps the value chain, distinguishes direct and indirect beneficiaries, and considers what may already be priced in. Its output is a candidate shortlist with methodology, comparison data, risks, catalysts, and research priorities; candidates are not investment conclusions.
Collects long/short direction, market-cap range, sector, style, geography, and thematic parameters; organizes stock screens using value, growth, quality, short, and special-situation criteria; researches thematic opportunities by defining the thesis, mapping the value chain, distinguishing pure-play from diversified exposure, and assessing priced-in expectations; presents each candidate with company, direction, one-line thesis, valuation and operating metrics, thesis points, risks, catalysts, and suggested next steps; produces a five-to-ten-idea shortlist, documented screening methodology, comparison table, and prioritized research list.
- An equity analyst needs new long or short candidates within a defined sector and starts with market-cap, style, and valuation filters.
- An investment team is researching AI infrastructure, reshoring, or aging demographics and wants to identify direct, indirect, and second-order beneficiaries.
- A researcher is looking for a quality company temporarily mispriced because of a headwind and combines quality and value screens.
- A short seller wants a structured review of slowing revenue, margin compression, working-capital anomalies, insider selling, and accounting risks.
- A research team is building an idea pipeline and wants every candidate documented with a catalyst, key risks, and a next-step recommendation.
Pros & cons
- Provides frameworks for value, growth, quality, short, and special-situation screens.
- Combines quantitative criteria, thematic value-chain analysis, and pattern recognition.
- Defines a concrete candidate format with metrics, thesis, risks, catalysts, and next steps.
- Encourages research discipline around catalysts, crowded trades, analyst coverage, and historical hit rates.
- The SKILL.md does not identify data sources, real-time market feeds, or a screening tool, so data acquisition is unspecified.
- It surfaces candidates rather than delivering a full model, valuation conclusion, or trade execution.
- The source provides no test suite, platform validation record, or concrete output example.
- The README states that the repository is not investment advice and that outputs require qualified professional review and compliance checks.
How to install
The skill is located at plugins/vertical-plugins/equity-research/skills/idea-generation/SKILL.md. According to the repository README, add the marketplace with claude plugin marketplace add anthropics/financial-services, then install the vertical plugin with claude plugin install equity-research@claude-for-financial-services. In Cowork, open Settings → Plugins → Add plugin, paste https://github.com/anthropics/financial-services, and select the relevant vertical. The source does not document installation of this individual skill separately.
How to use
After installing the equity-research vertical plugin, use the /screen command listed in the README or send a request matching the documented triggers, such as “screen for large-cap US quality stocks” or “find ideas in AI infrastructure.” Provide direction, market cap, sector, style, geography, and theme when applicable. The skill then structures the screening and requires further fundamental work on candidates. The source does not specify a particular data provider, API, or automated execution method.