Quick Stock Analysis

A fast, opinionated screen that answers one question: should I bother doing deeper fundamental analysis on this stock? It runs through 5 gates. All five gates are always evaluated — a failing gate never ends the screen — and the outcome is reported as a score out of five (e.g. 4/5) alongside the per-gate breakdown, because a stock that misses one gate can still be interesting. Only a 5/5 earns an "Invest" verdict.

The stock symbol MUST be provided by the user. If not, ask for it.

This is a screening tool, not an investment decision. A passing verdict means "worth a closer look" — always follow up with the full-stock-analysis skill before committing capital.

Data Access

All data comes from the FMP MCP connector. Do not use WebSearch, a scraped page, or a local library for any gate value — if FMP cannot supply a field, the field is UNKNOWN. Invoke the market-data skill if you need the exact endpoint or field name for anything below.

Make these four calls before evaluating any gate:

1. statements(endpoint="income-statement",        symbol=<TICKER>, period="annual", limit=5)
2. statements(endpoint="metrics-ratios",          symbol=<TICKER>, period="annual", limit=5)
3. statements(endpoint="balance-sheet-statement", symbol=<TICKER>, period="annual", limit=5)
4. quote(endpoint="quote", symbol=<TICKER>)

Two optional calls, used only where noted:

5. company(endpoint="profile-symbol",       symbol=<TICKER>)   # industry + description for frontmatter
6. analyst(endpoint="financial-estimates",  symbol=<TICKER>, period="annual")   # forward EPS growth for PEG

Rows come back newest-first. If call 1, 2, or 4 fails or returns an empty array, report the error to the user and stop — the screen is not meaningful without them.

If the ticker is ambiguous or unrecognised, resolve it first with search(endpoint="search-symbol", query=<user input>) and confirm the match with the user before screening.

Deriving the Gate Values

FMP returns ratios as fractions (0.147 = 14.7%). Convert before comparing against a percentage threshold, and display them as percentages.

Decision Tree

Evaluate all five gates on every run, in order. Never short-circuit on a failure: a stock that fails gate 1 but clears the other four is a very different proposition from one that fails everything, and only a complete scorecard shows the difference.

Gate 1 — Revenue Growth

• Question: Has revenue grown by at least 10% every year for the last 3–5 years?
• Metric: YoY revenue growth rate (annual)
• Pass: Every year ≥ 10%
• Fail reason: Low revenue growth

Gate 2 — P/E Ratio

• Question: Is the trailing P/E ratio below 25?
• Metric: Trailing twelve-month P/E
• Pass: P/E < 25 (and positive — negative P/E fails this gate)
• Fail reason: Likely overvalued

Gate 3 — PEG Ratio

• Question: Is the PEG ratio below 2?
• Metric: PEG = P/E ÷ EPS growth rate (%)
• Pass: 0 < PEG < 2
• Fail reason: Low profit growth

Gate 4 — Return on Equity

• Question: Has the company's ROE been over 5% on average for the last 5 years?
• Metric: 5-year average ROE
• Pass: Average ROE > 5%
• Fail reason: Weak profitability

Gate 5 — Quick Ratio

• Question: Is the quick ratio above 1.5?
• Metric: Quick Ratio = (Current Assets − Inventory) ÷ Current Liabilities
• Pass: Quick Ratio > 1.5
• Fail reason: Liquidity issues

Score and Final Verdict

Count the gates that PASS — UNKNOWN never counts as a pass — and report the total as <passed>/5.

If any gate is UNKNOWN, append INCONCLUSIVE to the verdict and say which gates could not be evaluated. A 4/5 with one UNKNOWN is not the same as a 4/5 with one FAIL — never present them as equivalent.

Output

Write the result to ./Stock Screener/<TICKER>.md in the current working directory. Use the template below. Fill every field — use N/A or UNKNOWN when data is genuinely missing. invest is true only at 5/5; score, positive-gates and verdict always carry the full picture, so a 4/5 stays discoverable even though invest is false.

If the Stock is pre-existing, update the Note!.

---
tags:
  - stock
  - quick-analysis
date: <YYYY-MM-DD>
current-price: <price>
industry: <industry>
invest: <true/false>
score: <passed>/5
positive-gates: <number of gates passed, 0-5>
verdict: <INVEST | WATCH | WATCH (weak) | REJECT>
---

# Quick Stock Analysis: <TICKER> (<Company Name>)

**Score: <passed>/5 — <VERDICT>**

<Quick description of the Stock / Equity>

## Decision Tree Results

| #   | Gate           | Threshold    | Value   | Result            |
| --- | -------------- | ------------ | ------- | ----------------- |
| 1   | Revenue Growth | ≥ 10% yearly | <value> | PASS/FAIL/UNKNOWN |
| 2   | P/E Ratio      | < 25         | <value> | PASS/FAIL/UNKNOWN |
| 3   | PEG Ratio      | < 2          | <value> | PASS/FAIL/UNKNOWN |
| 4   | ROE (5y avg)   | > 5%         | <value> | PASS/FAIL/UNKNOWN |
| 5   | Quick Ratio    | > 1.5        | <value> | PASS/FAIL/UNKNOWN |

**Score: <passed>/5** — failing gates: <list, or "none">.

## Supporting Data

### Revenue (last 5 years)

| Year | Revenue | YoY Growth |
| ---- | ------- | ---------- |
| ...  | ...     | ...        |

### Valuation

- Trailing P/E: <value>
- Forward P/E: <value>
- PEG: <value>
- EPS growth rate used for PEG: <value> (<forward estimate | trailing 5y CAGR>)

### Profitability

- ROE (current): <value>
- ROE (5y average): <value>
- Net margin: <value>

### Liquidity

- Current ratio: <value>
- Quick ratio: <value>
- Cash & equivalents: <value>

## Interpretation

<2-4 sentences explaining the score. Name every failing gate and what would need to change
for each, and call out the strongest metric. For 5/5 or 4/5, recommend running the
`full-stock-analysis` skill next. For 3/5, say whether the two misses look explainable.
For 0-2/5, say what to watch for a future revisit.>

## Next Step

- 5/5 or 4/5 → run `full-stock-analysis` on <TICKER> for DCF and peer comparison.
- 3/5 → optional deeper look; decide on the merits of <the failing gates>.
- 0-2/5 → skip deeper analysis; revisit if <the failing metrics> improve.

---

_Data: FMP connector, fiscal years <first>–<last>. Quote as of <date>._
_Quick screen only — not a buy/sell recommendation. Not financial advice._

Workflow

1. Confirm the ticker with the user if not provided.
2. Fetch the data — make the four required FMP calls above. This step is mandatory and must complete before any gate is evaluated.
3. Derive every gate value using the derivation table and evaluate all five gates. Do not stop early on a FAIL.
4. Count the passes and map the score to a verdict per the score table. Add INCONCLUSIVE if any gate is UNKNOWN.
5. Write the markdown file using the template above.
6. Print the score (<passed>/5), the verdict, the failing gates, and the file path in the chat response.

Anti-Patterns

• Don't skip gates because "the company is famous." The whole point is a mechanical screen.
• Don't short-circuit the decision tree. All five gates get calculated on every run, even when gate 1 fails outright — the one-gate miss is often the interesting case.
• Don't fudge missing data. If revenue growth for one year is missing, mark UNKNOWN — do not interpolate.
• Don't substitute a web search for a failed FMP call. A scraped number is not comparable to the rest of the table. Mark it UNKNOWN.
• Don't mix period-end and live figures silently. The quote P/E is live; metrics-ratios P/E is as of the fiscal year end. Say which you used.
• Don't treat INVEST as a buy signal. It's a "worth deeper research" flag, nothing more.
• Don't collapse the score into a bare pass/fail. 4/5 and 1/5 are both "not INVEST" and are not remotely the same thing; always show the score and which gates failed.
• Don't apply this to banks, insurers, REITs, or pre-profit growth stocks. Quick ratio and P/E are misleading for financials; ROE and P/E are misleading for unprofitable growth. If the user requests one of these, note the limitation and suggest full-stock-analysis directly instead.