US Stock Fundamentals & Screening with AI Agents
Skip the manual filing grind in value investing. Run the whole chain with an AI agent: screen candidates, verify TTM metrics and financial ratios, rank by price change, and output a shortlist from plain-language criteria.

Tool names in this guide are for reference only and may have changed or been renamed. Please use the latest tools in the tool marketplace when implementing.
Key takeaways
- Define screening criteria in plain language and run screen-to-shortlist in one chain
- After the initial screen, verify each stock with TTM metrics and financial ratios
- Rank the candidate pool by price change instead of opening filings one by one
Tools used
Why fundamental stock picking eats so much time
The logic of value investing is simple: find financially healthy companies at a reasonable price. The execution is not. You screen thousands of tickers by market cap, P/E, ROE and leverage, then open each filing to check TTM (trailing-twelve-month) metrics and financial ratios, and finally watch price moves to prioritize the list.
Every step lives on a different site, report or spreadsheet. By the time you've worked through a few dozen names by hand, the market may have moved — and it's easy to miss a criterion or mistype a number. What you really want is to hand the whole "set criteria → screen → verify → rank" chain to an agent.
The workflow in five steps
- Tell the agent your screening criteria in plain language (market-cap range, sector, valuation caps, etc.).
- Run an initial screen to get a pool of candidate tickers.
- Pull TTM metrics and financial ratios for each candidate to verify quality.
- Have the agent rank the pool using price-change data.
- Output a shortlist of candidates with their key metrics.
How each step works
Step 1: State your criteria
Just tell the agent your stock-picking logic, e.g. use ae to screen US companies with a market cap above $5B, P/E under 20, and ROE above 15%. Describe the conditions in natural language; the agent translates them into screener parameters.
Step 2: Screen for candidates
Use a screener-type tool (such as ae_fmp_get_stable_company_screener) to narrow the whole market to a candidate pool by market cap, sector and valuation. This replaces dragging sliders and paging through a screening site, returning a list of tickers that match your conditions.
Step 3: Verify TTM metrics and ratios
For each candidate, use a key-metrics tool (such as ae_fmp_get_key_metrics_ttm) to read trailing-twelve-month per-share metrics, valuation multiples and returns; then a ratios tool (such as ae_fmp_get_ratios_ttm) to check profitability, solvency and operating ratios. For example: use ae to get AAPL's TTM key metrics and financial ratios. The TTM basis avoids the distortion of looking at a single quarter.
Step 4: Rank by price change
Use a price-change tool (such as ae_fmp_get_stable_stock_price_change) to pull each candidate's moves over different windows, and let the agent rank the pool accordingly — for instance, surfacing names that pass on fundamentals but have pulled back recently.
Step 5: Produce the shortlist
Have the agent compile the matched, verified and ranked results into a single table, each row carrying its key metrics, ready to use.
Manual screening vs. agent-driven screening
| Stage | Manual screening | Agent-driven screening |
|---|---|---|
| Set criteria | Sliders across sites | One plain-language prompt |
| Initial screen | Page and export by hand | Screener returns candidates at once |
| Verify metrics | Open each filing | Batch-pull TTM metrics/ratios |
| Ranking | Build the table yourself | Agent ranks by price change |
| Reproducibility | Redo it next time | Change criteria and rerun |
FAQ
Q: Can I customize the screening criteria? Yes. Market cap, sector, valuation multiples, returns and more can be specified in natural language; the agent maps them to screener parameters and filters out non-matches.
Q: Are TTM (trailing-twelve-month) metrics available? Yes. Both key metrics and financial ratios can be pulled on a trailing-twelve-month basis, avoiding the distortion of a single quarter.
Q: Which markets are covered? Fundamentals for US and other globally listed companies. Exact coverage follows each tool's description in the marketplace.
Getting started
- Sign up — new users get 1,000 free credits.
- Connect AgentEarth to your agent — in-app search in WorkBuddy / Trae, or an MCP config line for Claude Code / Cursor / VSCode.
- Give instructions in natural language; the agent picks the tools and runs the full screening chain.
