01
Discover
Read the contract
Start with the Agent Card, OpenAPI or llms.txt. Do not scrape the human page to infer a request.
Evidence-backed financial answers for autonomous agents
Agents can discover a small set of stable research products, call them directly and receive bounded answers with coverage, freshness, tools used and lineage. Free and machine-payable queries share the same evidence contract.
Example outputs
Answers that carry their evidence.
These compact examples were captured from production on 1 September 2026. They show the response shape and evidence metadata; values remain historical rather than silently presenting themselves as live quotes.
Company snapshot · AAPL
Factual issuer context
Adjusted close$316.85
Market cap$4.67T
P/E ratio36.29×
Gross margin50.06%
- As of
- 2026-08-31
- Coverage
- 4/4 tools available
- Source
- fmp_bulk
- Lineage
- 3 request references
Earnings preview · AAPL
Consensus and revision evidence
Current EPS · 2026 Q31.98483
30-day drift−1.57%
Analyst coverage18
Review questions4
- As of
- 2026-08-31
- Coverage
- 9/9 tools available
- Freshness
- per-tool timestamps
- Lineage
- 8 request references
Machine flow
Discover, authorize, query, verify.
The machine contract is the product. It tells the caller what can be invoked, what it costs, which scope is required and what evidence fields will be returned before the agent spends time or money.
02
Authorize
Use the declared access mode
Call no-auth products directly. Register an OAuth client only when the chosen operation requires agents:query.
03
Execute
Send a bounded payload
Use the exact schema and limits from OpenAPI. Paid endpoints return the x402 challenge before execution.
04
Verify
Inspect evidence
Preserve coverage, as-of dates, freshness, tools used, warnings and lineage references with the answer.
Evidence contract
The answer explains its own limits.
factsinterpretationcoverageasOffreshnesstoolsUsedlineageRefswarnings
Try a free query
Give this to an HTTP-capable coding agent.
The free company snapshot needs no QJ account and no payment. A browser-only chat session may be able to read the contract but still lack permission to send outbound POST requests; coding agents and API clients can execute it directly.
What the caller receives
A compact factual artifact, not a recommendation.
The response separates facts from interpretation and reports missing coverage rather than filling gaps.
Use another ticker by changing only the bounded ticker field.
Discover QuantJourney at:
https://agents.quantjourney.cloud/openapi.json
Call POST /v1/query/company-snapshot-lite with:
{"ticker":"AAPL"}
Return the exact artifact, then summarize the facts,
freshness, missing coverage and lineage. Do not turn
the result into an investment recommendation.