Read before you trust a signal

Methodology

How 13FLOW turns public SEC filings into read-only research surfaces, how the validation boundary constrains claims, and how the MCP layer exposes that context to agents.

Application methodologyMCP methodologyConfluence v1 JSON

Application methodology

Data pipeline, 13F caveats, quality warnings, Confluence scope and validation boundary.

MCP methodology

Tool contract, Pro gating, fail-closed behavior, x402 readiness and agent safety.

Validation pack

Mechanical evidence, descriptive metrics, current artifact hashes and explicit non-claims.

How to use the tool

01 · ConfirmOpen Status and verify LIVE, commit SHA and latest 13F quarter.
02 · ScreenUse Cockpit, Signals, Funds and Stocks to build a research queue.
03 · InspectFollow SEC accessions, manager context, Form 4 overlap and quality warnings.
04 · DecideTake decisions outside 13FLOW, with your own model, risk limits and source review.

What 13FLOW is good at

  • Making delayed SEC 13F disclosures navigable.
  • Joining institutional movement with Form 4 evidence where available.
  • Surfacing methodology, source links and data-quality warnings beside the signal.
  • Giving humans and agents the same public contracts.

What users must remember

  • 13F filings are delayed regulatory disclosures, not live books.
  • Scores are ordinal research screens, not probabilities or expected returns.
  • Validation is mechanical and conservative until broader reviewed artifacts are published.
  • Redistribution and production Pro use require operator approval.