DFX Intelligence

Research one question in one call

research

Read-onlyIdempotent

One private markets question in one call: BDC and private credit lenders, borrowers, loans, maturities and non-accrual, sponsors, advisers, teams, pension commitments. Reads SEC filings (BDC schedules, Form ADV, IAPD) and plan documents; returns entities, an answer from cited claims only, coverage limits, withheld counts. identify and screen depth are free. Paged by continuation. No model inside.

Input Schema

Name	Required	Description	Default
`as_of`	No	Optional ISO date the answer is read as of (default: today).
`depth`	No	Optional. The default depends on access: without a paid DFX plan or the 7 day trial (anonymous, a free key or a free connection) only identify and screen run, and an unnamed depth runs at screen; with one, an unnamed depth runs at investigate. identify: ids only (up to 6 reads, 3 KB). screen: the primary read at the latest date (8 reads, 8 KB). investigate: every date and section the question needs (14 reads, 12 KB a page). validate: investigate plus cross-checks, disagreements under contradictions (18 reads). monitor: changes since `since` or 90 days, with a checkpoint (8 reads).
`since`	No	Optional ISO date: the window start (depth=monitor: the checkpoint).
`detail`	No	compact: identity, the answer, its basis claims and counts, under 2 KB. research (default): the page in full.	research
`entity`	No	Optional. The subject: a name (firm, fund, company, plan, advisor) or a DFX id (dfx:<graph>:<id>). Without it, every entity is read from the question.
`continue`	No	page.continuation from a previous answer: returns its next page, read through this caller's own access. Pages split between whole claims, never inside one.
`question`	No	The question in plain language, as an analyst would ask it.
`workflow`	No	Optional hint. credit_by_sponsor: BDC lenders, positions, facilities and sponsors. advisor_movement: IAPD registration moves, people and teams. allocator_to_company: plan commitments and capital paths. Without it, the workflow is read from the question's words.

Output Schema

Name	Required	Description
`ok`	No
`next`	No
`page`	No	index, rows and claims on this page and in all, rows after it, continuation
`plan`	No	workflow, operation, depth, the operations run and the budget
`dates`	No
`facts`	No
`state`	Yes	OK, NO_MATCH, NOT_COVERED, UNKNOWN, or an error state
`usage`	No	operations run, budget, bytes, latency; no model call
`answer`	No	text, basis (claim ids on this page), state
`claims`	No	every claim the rows on this page carry (lane I1's claims.research_claims): claim_id (content addressed: subject, predicate, object, value, date, source), row, predicate, claim_type, evidence_status, entitlement_class, derived_from; the subject, value, dates and citation are on the row it names
`events`	No
`schema`	Yes
`revisit`	No	how to come back to this answer: for a key whose plan includes watches, one watch_create call that saves the question; otherwise what that needs, plus meanwhile_call (a free search_private_credit_changes call on the subject since the answer's as_of) where this profile lists that tool. Private credit subjects only; absent when it does not apply
`question`	No
`withheld`	No
`identities`	No	the subject, every other entity resolved, candidates
`claims_meta`	No
`uncertainty`	No
`answer_state`	No
`premise_check`	No
`relationships`	No
`contradictions`	No
`coverage_limits`	No
`do_not_conclude`	No

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedOct 26, 2026

removedInput schema / properties / depth / default
```
Removed value: -"investigate"
```

changedInput schema / properties / depth / description

Previous value: -"identify: ids only (up to 6 reads, 3 KB). screen: the primary read at the latest date (8 reads, 8 KB). investigate (default): every date and section the question needs (14 reads, 12 KB a page). validate: investigate plus cross-checks, disagreements under contradictions (18 reads). monitor: changes since `since` or 90 days, with a checkpoint (8 reads). Without a paid DFX plan or the 7 day trial, identify and screen run and an unnamed depth runs at screen."New value: +"Optional. The default depends on access: without a paid DFX plan or the 7 day trial (anonymous, a free key or a free connection) only identify and screen run, and an unnamed depth runs at screen; with one, an unnamed depth runs at investigate. identify: ids only (up to 6 reads, 3 KB). screen: the primary read at the latest date (8 reads, 8 KB). investigate: every date and section the question needs (14 reads, 12 KB a page). validate: investigate plus cross-checks, disagreements under contradictions (18 reads). monitor: changes since `since` or 90 days, with a checkpoint (8 reads)."

Changed1 schema field changedOct 26, 2026

addedOutput schema / properties / revisit

Added value: +{
+  "description": "how to come back to this answer: for a key whose plan includes watches, one watch_create call that saves the question; otherwise what that needs, plus meanwhile_call (a free search_private_credit_changes call on the subject since the answer's as_of) where this profile lists that tool. Private credit subjects only; absent when it does not apply",
+  "type": "object"
+}

Changed2 schema fields changedOct 25, 2026

changedInput schema / properties / continue / description

Previous value: -"page.continuation from a previous answer: returns its next page."New value: +"page.continuation from a previous answer: returns its next page, read through this caller's own access. Pages split between whole claims, never inside one."

changedInput schema / properties / depth / description

Previous value: -"identify: ids only (up to 6 reads, 3 KB). screen: the primary read at the latest date (8 reads, 8 KB). investigate (default): every date and section the question needs (14 reads, 12 KB a page). validate: investigate plus cross-checks, disagreements under contradictions (18 reads). monitor: changes since `since` or 90 days, with a checkpoint (8 reads)."New value: +"identify: ids only (up to 6 reads, 3 KB). screen: the primary read at the latest date (8 reads, 8 KB). investigate (default): every date and section the question needs (14 reads, 12 KB a page). validate: investigate plus cross-checks, disagreements under contradictions (18 reads). monitor: changes since `since` or 90 days, with a checkpoint (8 reads). Without a paid DFX plan or the 7 day trial, identify and screen run and an unnamed depth runs at screen."

AddedOct 22, 2026

TDQS

A4/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already cover read-only/idempotent/non-destructive/open-world, but the description adds substantive behavior: source documents read, answer constrained to cited claims, coverage limits and withheld counts returned, free depth levels, continuation paging, and no model inside. This is exactly the extra context annotations cannot supply.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded paragraph with no filler sentences. The long domain enumeration is dense but still purposeful; minor cost is that the list of entity types runs together.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a complex research tool with eight optional params, an output schema, and rich annotations, the description supplies the essential mental model: sources, answer grounding, coverage limits, free tiers, paging, and no model. The remaining gap is sibling routing, which overlaps with usage guidance.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already documents all eight parameters including enums. The description adds access/cost context for the depth choices and the continuation paging behavior, though most detailed parameter semantics remain in the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific operation: answer one private markets question in one call by reading SEC filings and plan documents, returning entities plus a cited answer, coverage limits, and withheld counts. That is far beyond restating the name, though it does not explicitly name the granular search_* siblings it replaces.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives operating context (one question per call, identify/screen free, paged via continuation) but never states when to choose this over the many search_* and get_* siblings or when not to use it. The access constraint is useful but is not alternative routing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All DFX Intelligence tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.