Choose Your Surface¶
Disclosure Alpha exposes the same deterministic pipeline through a CLI, Python SDK, focused HTTP endpoints, and two MCP bundles — pick the surface that matches your workflow.
Which surface?¶
flowchart TD
start["What do you have?"]
ticker["Ticker + fiscal year"]
html["Raw filing HTML"]
batch["Many tickers"]
agent["AI agent / MCP client"]
start --> ticker
start --> html
start --> batch
start --> agent
ticker --> cli_t["CLI or Python SDK"]
ticker --> http_t["HTTP matrix GET"]
ticker --> mcp_a["MCP Analyst"]
html --> cli_h["CLI --html"]
html --> py_h["Python score_filing_html"]
html --> mcp_b["MCP Builder"]
batch --> panel["HTTP Panel POST"]
cli_t --> read["Read scores: understanding-scores"]
http_t --> read
panel --> read
Need help reading the JSON? Start with Understanding Scores.
Personas¶
Persona |
Need |
Surface |
|---|---|---|
Discovery / data engineer |
List filings, extract sections |
Filing Index + Section Extractor (HTTP) |
Quant / researcher |
Raw metrics, flags, diffs |
Disclosure Analytics, Flags, Changes (HTTP) |
Risk analyst |
Filing-level scores |
Disclosure Risk Score / matrix (HTTP) |
Screener / index builder |
Batch tickers |
Panel POST (HTTP) |
Agent builder |
Low-level pipeline in MCP |
Builder MCP bundle |
Agent user |
Ticker score + filings |
Analyst MCP bundle |
Script / notebook |
One-off scoring |
CLI or Python SDK |
Quick comparison¶
Surface |
Best for |
Entry point |
|---|---|---|
CLI |
Terminal workflows, local HTML |
|
Python SDK |
Custom pipelines, notebooks |
|
HTTP API |
Services, dashboards, screeners |
|
MCP Analyst |
Cursor / Claude ticker tools |
|
MCP Builder |
Raw HTML agent workflows |
|
MCP bundles¶
Entry point |
Tools |
Resource |
|---|---|---|
|
|
|
|
|
— |
Legacy disclosure-alpha-mcp aliases to the analyst bundle.
Related¶
Understanding Scores — interpret score JSON
HTTP API Guide — endpoint map, tiers, Postman collections