Financial Modeling Suite
A financial modeling toolkit for valuation, risk assessment, and investment decisions.
The code appears to process user-supplied financial data in memory and shows no network access, credential collection, or destructive external action; the repository provides MIT licensing and publisher provenance, and SKILL.md includes financial disclaimers. Points are deducted because data flows, sensitive-financial-data handling, user-confirmation boundaries, and rollback are not specified, while default assumptions and controls for decision risk are insufficiently explicit.
The DCF example covers a plausible basic path through historical data, assumptions, WACC, cash flows, enterprise value, and equity value, and dependencies are visible in pyproject.toml. Points are deducted because there are no tests for this skill; validation is missing for length mismatches, zero revenue, invalid WACC/terminal-growth relationships, and parameter ranges. Sensitivity analysis can silently accept unknown variables, scenario_analysis does not fully restore model state, and some failures shift debugging to the user. The static-review ceiling also limits this score.
The main inputs, DCF/sensitivity/scenario use cases, and example requests are reasonably clear. Points are deducted because SKILL.md claims Monte Carlo and broader outputs while the supplied scripts only implement DCF and sensitivity analysis; non-fit boundaries, trigger conditions, delivery workflow, and Chinese-language support are not defined, weakening semantic invocation accuracy.
The documentation is layered across capabilities, inputs, outputs, model types, practices, and limitations. The repository also provides an MIT license, version, Python requirement, dependencies, and official organizational provenance. Points are deducted for sparse installation and calling instructions, no skill-specific version history or changelog, no clearly assigned maintainer or update path, and unsupported claims about automatic quality checks, market-parameter updates, and continuous improvement.
The included examples can produce a basic DCF and sensitivity-analysis table, with relatively direct result structures. Points are deducted because the claimed Monte Carlo simulation, scenario probability modeling, charts, Excel workbook, and complete model outputs are not implemented; the financial model uses material simplifications and requires substantial professional review, with no verified representative outputs. Static review caps this dimension at 7.
The selected path includes auditable Python source, documentation examples, and repository-level dependency/CI configuration, so it is more than unverifiable marketing. Points are deducted because there is no committed test coverage or third-party execution evidence for this path; the CI workflows mainly review links and model changes and do not establish reproducibility of the key paths. Static review therefore caps this dimension at 5.
- The claimed Monte Carlo, scenario-planning, Excel, and chart outputs are not implemented in the supplied scripts.
- The perpetuity-growth terminal-value path does not validate that WACC exceeds terminal growth, risking invalid or extreme valuations.
- scenario_analysis, create_data_table, and parts of the sensitivity workflow may not fully restore model state; results should not be used directly for investment decisions.
- Storage, transmission, retention, and access controls for sensitive financial data are unspecified, and no Chinese-language usage guidance is provided.
What it does & when to use it
This skill supports investment analysis, valuation, and risk assessment through DCF modeling, sensitivity analysis, Monte Carlo simulation, and scenario planning. It covers corporate valuation, project finance, M&A analysis, and LBO models. Inputs include historical financials, forecast assumptions, risk parameters, probability distributions, and scenario weights; outputs include valuation results, sensitivity tables, statistical summaries, and risk metrics. Results depend on the assumptions and are not a substitute for professional financial advice.
Reads historical financial statements, growth assumptions, margin projections, capital expenditure, working capital, terminal value inputs, discount-rate components, variable ranges, probability distributions, correlations, iteration counts, and scenario weights; builds DCF projections and calculates terminal value, WACC, enterprise value, and equity value; runs sensitivity tests, tornado charts, Monte Carlo simulations, and best/base/worst-case comparisons; produces projections, free-cash-flow calculations, valuation ranges, confidence intervals, statistical summaries, VaR, probability-weighted outcomes, and Excel model output; includes dcf_model.py and sensitivity_analysis.py.
- An investment analyst imports financials for a technology company and builds a DCF with multiple growth scenarios.
- An M&A team runs 5,000 Monte Carlo iterations on an acquisition model to evaluate valuation distributions and target-achievement probabilities.
- A valuation professional tests how growth rate and WACC affect valuation and produces sensitivity tables and a tornado chart.
- A project team develops probability-weighted best-case, base-case, and worst-case scenarios for an expansion project.
Pros & cons
- Covers DCF, sensitivity analysis, Monte Carlo simulation, and scenario planning.
- Supports corporate valuation, project finance, M&A, and LBO model types.
- Provides valuation, statistical, risk, and chart-oriented outputs.
- Lists quality checks for balance sheets, cash flows, sensitivity bounds, and Monte Carlo results.
- The source provides no standalone installation command, runbook, or test-suite information.
- Model quality depends on assumptions, and market, regulatory, and tax changes can affect results.
- Professional judgment is required; the tool is not a substitute for professional financial advice.
- The source does not specify supported Python versions, platforms, or Excel-generation details.
How to install
The skill is located at skills/custom_skills/creating-financial-models/SKILL.md in the GitHub repository anthropics/claude-cookbooks. The source does not provide a standalone installation command or client-specific placement instructions. The repository README says its examples generally require a Claude API key.
How to use
Provide financial statements and relevant assumptions, then use prompts such as “Build a DCF model for this technology company using the attached financials” or “Create sensitivity analysis showing impact of growth rate and WACC on valuation.” The source does not document script invocation, client configuration, or parameter passing.