Can I send a natural-language or SQL screener?

No. The input is a strict typed filter array with a fixed field and operator registry.

Is an empty match list an error?

No. A complete scan with zero matches is a valid result and preserves full universe and field-coverage counts.

Can AI predict stock prices with this MCP tool?

No. AI cannot predict stock prices with certainty, and this MCP tool does not promise future returns. It gives an agent bounded data or validation results that still require independent review, risk controls, and source checking.

Agent workflows

Stock Screening MCP Tool for AI Agents

screen-stocks applies one to eight ordered AND filters to a canonical S&P 500, Nasdaq-100, or Dow 30 universe. It supports sixteen numeric fields plus GICS sector, one validated sort field, a result limit of 1–25, missing-value failure semantics, and per-match observed evidence.

Connect Pineify MCP View all MCP tools

MCP tool name

screen-stocks

Example prompt

Use Pineify screen-stocks on the S&P 500 for market cap at least $1B and P/E below 30, returning five matches.

Direct answer

What screen-stocks does

The input is a strict typed filter array, not free text, SQL, code, or a screener DSL. Duplicate fields, invalid operators, reversed bounds, and unsupported sort fields fail before provider calls.

Fundamental and 1d technical facts are joined by canonical symbol. Every returned match is re-evaluated against observed facts, with coverage and truncation counts preserved.

Contract

What the agent sends and receives

Inputs

`universe`: sp500, nasdaq100, or dow30.
`filters`: one to eight typed numeric/between/sector filters evaluated as AND.
`sort`, `limit` from 1 to 25, and optional compact or standard `detail`.

Structured output

`stock-screening-v1` structured content.
Canonical universe metadata and requested, deduped, covered, evaluated, matched, returned, and truncated counts.
At most 25 matches with observed facts, ordered filter evidence, coverage, warnings, and source times.

Capabilities

Where this MCP tool fits

Use canonical universes

Deduplicates versioned S&P 500, Nasdaq-100, or Dow 30 constituents and caps evaluation at 500 members.

Combine typed facts

Joins profile, valuation, growth, sector, RSI, moving-average distance, and relative-volume fields.

Return auditable matches

Provides per-filter evidence, field coverage, null-last sorting, stable symbol ties, and matched/returned counts.

Agent workflow

A bounded call from question to review

1
Build typed filters
Choose fields, compatible operators, and bounded numeric or sector values.
2
Audit universe coverage
Check deduped and covered counts before interpreting matches.
3
Verify every match
Read observed facts and evidence in the same order as the input filters.

Prompt examples

Questions an agent can route to this tool

"Use Pineify screen-stocks on the S&P 500 for market cap at least $1B and P/E below 30, returning five matches."

"Screen the Nasdaq-100 for RSI between 40 and 60 and sort by relative volume descending."

Operator notes

How I review the result

I distinguish a valid zero-match result from unavailable universe or provider coverage.

I do not treat missing values as passing and never infer an unobserved fundamental or technical value.

Boundaries to keep in the prompt

The tool supports only the fixed field registry, three universes, AND logic, and 1d technical facts.
Missing facts fail their filters; partial provider batches can reduce coverage.
Screen results are research candidates, not recommendations or executable orders.

Pineify MCP is an information and code-validation tool, not investment advice. It does not promise returns, place trades, or modify a portfolio. Review timestamps, source fields, code diagnostics, and risk assumptions before acting.

MCP tool family