What does this skill do, and when should you use it?
This skill supports first-time equity coverage through five tasks: company research, financial modeling, valuation analysis, chart generation, and report assembly. Each request must execute exactly one task, with prerequisite outputs verified before dependent tasks begin. Deliverables can include research markdown, an Excel model, valuation analysis, a chart archive, or a final DOCX report. The workflow is structured and detailed, but it requires financial data, external market data, and document or spreadsheet capabilities for the later stages.
It reads a company name or ticker, financial information, and outputs from earlier tasks; performs qualitative research, historical financial extraction, five-year projections, scenario analysis, DCF valuation, and comparable-company analysis; generates 25–35 financial charts; and assembles the research, model, valuation, and charts into a final DOCX report. Task 3 depends on Task 2, Task 4 depends on Tasks 1–3 plus external market data, and Task 5 depends on all prior tasks.
- An equity researcher starting coverage of a company can run Task 1 for business, management, industry, competition, TAM, and risks.
- A financial analyst with a 10-K or financial statements can run Task 2 to build the six-tab projection model.
- A research team with a completed model can run Task 3 for DCF, trading comparables, valuation ranges, and a price target.
- A report producer with Tasks 1–3 complete can run Task 4 to generate and package professional charts.
- A team with all four prior deliverables can run Task 5 to assemble the final 30–50-page DOCX report.
How do you install this skill?
- Before execution, confirm that the referenced files, DOCX/XLSX capabilities, and external data sources are actually available, and define recovery or rollback behavior for failures.
- Reconcile the six-tab model specification with the checklist requirement for 15 or more tabs, and resolve the Python-versus-DOCX workflow conflict.
- The report may contain company financial, management, and market data; data sourcing, freshness, access rights, privacy handling, and human fact-checking responsibility should be explicit.
- Reachability and authorization for SEC, Yahoo Finance, Bloomberg, and similar sources are not verified and may limit usability for mainland-China users.
- Network access
- Local filesystem
DOCX skillXLSX skill
The source does not document a standalone installation for this skill. In Cowork, open Settings → Plugins → Add plugin, paste https://github.com/anthropics/financial-services, and select the equity-research vertical plugin; alternatively upload a zip containing plugins/vertical-plugins/equity-research/. For Claude Code, the documented commands are: claude plugin marketplace add anthropics/financial-services; claude plugin install equity-research@claude-for-financial-services.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/anthropics/financial-services.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/plugins/vertical-plugins/equity-research/skills/initiating-coverage" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Use initiating-coverage skill, Task 1 for Tesla
Use the /initiate command listed in the README, or explicitly request one task, for example: “Use initiating-coverage skill, Task 1 for Tesla”. Execute only one task per request; when asked for the full pipeline, ask to begin with Task 1. Task 2 requires financial data, Task 3 requires the Task 2 model, Task 4 requires Tasks 1–3 and external market data, and Task 5 requires all prior deliverables.
What are this skill's strengths and limitations?
- Defines clear task boundaries and exact deliverables for staged review.
- Covers the full workflow from research and modeling through valuation, charts, and DOCX assembly.
- Includes explicit prerequisite checks and specifies required valuation and chart components.
- Supports Bull/Base/Bear scenarios, DCF analysis, and comparable-company analysis.
- It cannot run the five tasks automatically in sequence; users must make separate requests and advance the workflow manually.
- Tasks 2–5 depend on financial files, external market data, or earlier outputs, and the instructions prohibit placeholder work when inputs are missing.
- The source provides no standalone test suite or concrete connector configuration for the skill.
- Task 5 references DOCX and XLSX skills, but those capabilities are not defined inside this SKILL.md.
- The README states that outputs are analyst work product for professional review, not investment, legal, tax, or accounting advice.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Equity Research Initiation Workflow this page ✓ Anthropic · Official | 44 · Not recommended | ★ 40k | 1mo ago | Apache-2.0 |
| Earnings Update Researcher ✓ Anthropic · Official | 54 · Use with care | ★ 40k | 1mo ago | Apache-2.0 |
| StockValuation.io Local Valuation Workflow | 58 · Recommended | ★ 33 | 1mo ago | Apache-2.0 |
| Financial Model Refresh ✓ Anthropic · Official | 47 · Use with care | ★ 40k | 1mo ago | Apache-2.0 |
| Taiwan Stock One-Click Financial Model & Investment Deck Generator | 45 · Not recommended | ★ 40 | 3mo ago | — |
How did FollowSkills review this skill?
The skill requires one-task-at-a-time execution, prerequisite checks, and user confirmation, reducing unauthorized chaining and excess output. However, it does not adequately disclose permissions for external data retrieval, file operations, third-party tools, sensitive financial-data handling, source attribution, or rollback after failure, so substantial points are deducted.
Task order, prerequisites, and deliverables are described clearly, and missing inputs should cause the workflow to stop. However, executable availability is uncertain, referenced files and tool dependencies are not guaranteed by this file, and the template conflicts with the main document on Python versus DOCX workflow. Static review cannot reproduce key paths, so the score remains below the static ceiling.
The audience, five task types, inputs, and outputs are mostly clear, and single-task invocation is specified. Non-fit cases, semantic trigger boundaries, Chinese-language output support, and mainland-China environment fit are not defined; core research also depends on external sources such as SEC, Yahoo Finance, and Bloomberg, limiting demonstrated adaptability.
The documentation has layered task sections and includes naming rules, examples, a quality checklist, and several limitation disclosures. Versioning, changelog, maintenance ownership, update path, dependency installation, and troubleshooting are incomplete, and the materials conflict structurally, including six core Excel tabs versus a checklist requiring 15 or more tabs.
The skill specifies concrete deliverables for research, modeling, valuation, charts, and report assembly, so the core objective is plausibly achievable. However, the supplied files contain no real outputs, independently reviewable results, or third-party execution evidence, while the extensive page, word, chart, and table requirements may impose high cost and still require substantial manual review; the static cap supports 7.
The workflow includes detailed steps, checklists, source categories, and cross-file consistency checks, providing some auditability. It contains no committed test suite, CI coverage, representative output, or independent reproduction evidence, and its external sources are not fixed within the supplied skill materials, so only limited credit is warranted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →