mirror of
https://github.com/magnus919/agent-skills.git
synced 2026-09-14 21:16:36 +03:00
a45952d9c1
* feat(skill): beef up financial-modeling with templates/scripts/evals Add a schema-valid eval manifest (6 cases: unit-economics review, pricing decision, fundraising scenario, SaaS metrics interpretation, model sanity check, runway and burn analysis), four fillable templates (unit-economics record, pricing decision record, fundraising scenario, model sanity checklist), a stdlib SaaS-metrics calculator (ARR, monthly and annualized logo churn, NDR, Rule of 40) with a unittest suite, and a README Quick Start documenting the script. Closes #241. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com> * fix(skill): guard churn prints in saas-metrics human output print_human indexed monthly_logo_churn_pct and annualized_logo_churn_pct unconditionally while compute_metrics only populates them when churn inputs are given, so human-readable runs without churn inputs (--mrr alone, --mrr + NDR, --mrr + growth/margin) crashed with a KeyError (exit 1), violating the script's documented 0/2 exit-code contract. Guard both churn print lines with `if 'monthly_logo_churn_pct' in metrics:`, mirroring the existing NDR and Rule-of-40 guards, and add a regression test class covering human output with churn omitted. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com> --------- Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
64 lines
4.0 KiB
Markdown
64 lines
4.0 KiB
Markdown
# Financial Modeling
|
|
|
|
Build clearer financial scenarios, pricing decisions, and operating-metric analyses from stated assumptions.
|
|
|
|
## Why Install This Skill
|
|
|
|
Business decisions often depend on numbers that look precise but rest on hidden assumptions. This skill gives your agent a practical method for making revenue, costs, cash, customer economics, and SaaS metrics explicit so you can test what changes when the assumptions change.
|
|
|
|
It covers the questions teams repeatedly face: whether acquisition pays back, what a price change could do to retention, how long cash may last under different cases, and how to prepare a financing model or cap-table discussion. The material emphasizes traceable calculations and context rather than universal scorecards.
|
|
|
|
It is an analytical aid, not financial, investment, tax, accounting, or legal advice. Use qualified professionals where those disciplines are required.
|
|
|
|
## What You Get
|
|
|
|
| Path | What it provides |
|
|
|---|---|
|
|
| `SKILL.md` | Scope, trigger boundaries, and a concise assumptions-led working method. |
|
|
| `references/financial-modeling.md` | Linked statements, revenue and cost drivers, scenarios, and runway analysis. |
|
|
| `references/unit-economics.md` | CAC, LTV, payback, contribution margin, and segmentation guidance. |
|
|
| `references/pricing-strategy.md` | Pricing models, packaging, price-change tests, and trade-off analysis. |
|
|
| `references/fundraising.md` | Valuation methods, cap tables, term-sheet concepts, and fundraising preparation. |
|
|
| `references/saas-metrics.md` | ARR, retention, efficiency metrics, period alignment, and metric limitations. |
|
|
| `references/source-index.md` | Source provenance, review date, and authoritative reference links. |
|
|
| `templates/` | Fillable records: unit-economics record, pricing decision record, fundraising scenario, and model sanity checklist. |
|
|
| `scripts/` | `saas-metrics.py` — computes ARR, monthly and annualized logo churn, NDR, and Rule of 40 from stated inputs. |
|
|
| `evals/` | Output-quality eval manifest for the skill's methodology cases. |
|
|
|
|
## Quick Start
|
|
|
|
Compute the headline SaaS operating metrics in one command. The script needs only Python 3 (standard library) and takes stated inputs, so the same numbers you would put in a spreadsheet produce a consistent result:
|
|
|
|
```bash
|
|
python3 financial-modeling/scripts/saas-metrics.py \
|
|
--mrr 120000 --customers 480 --churned-customers 10 \
|
|
--expansion 9000 --contraction 3000 --churned-mrr 4200 \
|
|
--growth-pct 38 --margin-pct 6
|
|
```
|
|
|
|
Output:
|
|
|
|
```text
|
|
ARR (annualized recurring revenue): $1,440,000.00
|
|
Monthly logo churn: 2.08%
|
|
Annualized logo churn: 22.33%
|
|
NDR (net dollar retention): 101.50%
|
|
Rule of 40 (growth + margin): 44.00
|
|
```
|
|
|
|
Add `--json` for machine-readable output, pass `--churn-pct 2.1` instead of customer counts when churn is already known, and omit the `--expansion`/`--contraction`/`--churned-mrr` group (or the growth/margin pair) when those metrics are not in scope. The script exits 2 on inconsistent inputs, so it can gate a report or CI step. Load `SKILL.md` for the methodology and reference table, then use the templates to record unit economics, pricing decisions, fundraising scenarios, or a model sanity check.
|
|
|
|
## Triggers
|
|
|
|
- Build or review a financial model, forecast, budget, scenario, or runway analysis.
|
|
- Calculate CAC, LTV, payback, contribution margin, ARR, churn, NDR, Rule of 40, Magic Number, or burn multiple.
|
|
- Evaluate a pricing change, packaging model, fundraising scenario, cap table, valuation method, or term-sheet concept.
|
|
|
|
## Requirements
|
|
|
|
No API keys or credentials. The `saas-metrics.py` script needs only Python 3 (standard library). Useful analysis requires reliable business inputs with a defined currency, time period, and accounting basis.
|
|
|
|
## Source and Maintenance
|
|
|
|
This skill was ported from [`magnus919/hermes-profiles`](https://github.com/magnus919/hermes-profiles) at commit [`867a555`](https://github.com/magnus919/hermes-profiles/commit/867a555). See [`references/source-index.md`](references/source-index.md) for the portability boundary and source review details.
|