GET /analyses/:analysisId

Endpoint

GET /v1/analyses/{analysisId}

Request

curl -X GET "https://client-omni-api.argosidentity.com/v1/analyses/{analysisId}" \
  -H "x-api-key: your-api-key-here"

Response structure — two distinct groups

The first thing to establish when reading an analysis response is which fields the system returns and which fields you defined yourself.

System fields

Returned under the same keys regardless of workflow configuration. Execution history, per-step results, and aggregate metrics live here.

Output schema fields

outputSchema, extractedData, and extractionStatus. Their internal structure differs completely from workflow to workflow.

What appears inside extractedData is determined by the output schema you defined on your workflow. Omni does not impose a fixed set of fields. This page documents system fields field by field, and describes the output schema area structurally only.

System fields

id                profileId          folderId          engineId
playbookId        engine             status            processingTimeMs
verificationStatus  confidenceScore  confidence        details
riskAssessment    systemMetadata     rawActionResults  agentAuditLog
findings          recommendations    targetItems       error
options           clientMetadata     primaryReportId   reportId
requestedAt       completedAt        createdAt

Output schema fields

outputSchema      extractedData      extractionStatus

Top-level fields

string

required

Unique analysis ID (analysis_ prefix)

string

required

ID of the profile this analysis belongs to

string | null

required

Target folder ID. null when the whole profile was analyzed without specifying a folder

string | null

required

Engine ID when a single engine was explicitly specified. null when the workflow’s own engine configuration was used

object | null

required

Engine details when engineId is set. Otherwise null

string

required

Playbook ID used for the analysis (PB- prefix). Points to the execution plan generated from the workflow’s policy text

string

required

Processing state of the analysis job: pending / processing / completed / failed

number

required

Analysis processing time in milliseconds

string

required

System verdict for the analysis: verified / pending_review / rejected. See Three kinds of verification status below

number | null

required

Confidence produced alongside the verification verdict (0.0 – 1.0). For verified it is the mean confidence of the passed actions; for other verdicts it is a fixed value per verdict. A different metric from confidence

object | null

required

Error information when the analysis fails, carrying code, message, and optionally details. null on successful completion

object | null

Options passed when the analysis was requested. null if none were passed

object | null

Metadata the client passed when requesting the analysis. null if none was passed

string | null

ID of the primary report generated for this analysis (rpt_ prefix)

string | null

Legacy field holding the same value as primaryReportId. Use primaryReportId for new integrations

string

required

Analysis request time (ISO 8601)

string | null

required

Analysis completion time (ISO 8601). null while incomplete

string

required

DB record creation time (ISO 8601)

Three kinds of verification status

The response contains verification status at three layers, and the value sets differ. Always confirm which layer you are reading.

Location	Values	Nature
Top-level `verificationStatus`	`verified` / `pending_review` / `rejected`	System verdict for the whole analysis
`agentAuditLog.steps[].status` `agentAuditLog.summary.overall_decision` `rawActionResults.{actionName}.verificationStatus` `systemMetadata.workflowHistory[].verificationStatus`	`passed` / `needs_review` / `failed`	System execution result per playbook action (`overall_decision` is the aggregate of these)
Status fields inside `extractedData.{...}`	Whatever you defined in your schema	User-defined output — the system does not fix the value set

The second row is a playbook action execution result, not output schema data. If your output schema defines a field with the same name, it belongs to the third row and is separate from the system verdict.

The top-level verdict is decided by aggregating per-action results: rejected if any action failed, verified if all passed, and pending_review otherwise. agentAuditLog.summary.overall_decision is derived from the same action results, so the two correspond as follows.

`agentAuditLog.summary.overall_decision`	Top-level `verificationStatus`
`passed`	`verified`
`needs_review`	`pending_review`
`failed`	`rejected`

`confidenceScore` vs `confidence` — different metrics

The names are similar but they measure entirely different things.

	`confidenceScore`	`confidence` (Analysis Score)
Type	`number` (0.0 – 1.0)	`object`
Measures	Confidence tied to the verification verdict	Completeness of the analysis output (Analysis Score)
Derived from	Mean confidence of passed actions for `verified`; a fixed value for other verdicts	Step execution 40% + step quality 20% + output completeness 40%

A high confidenceScore does not imply a high confidence.score. The classic case: the AI confidently passed every step but not a single value landed in the output schema. If you build auto-approval logic, check confidence and details as well.

`confidence` — Analysis Score

This field is the Analysis Score shown at the top of the analysis detail in the Omni dashboard. The three components are combined into a weighted average to produce score, and level is the band that score falls into.

Field	Type	Description
`components`	`object`	The three inputs to the score. An input that cannot be measured is `null`
`score`	`number \| null`	Weighted average of the three inputs (0–100, rounded)
`level`	`string \| null`	Band `score` falls into: `HIGH` (80–100) / `MEDIUM` (65–79) / `LOW` (0–64)

`confidence.components`

Field	Weight	Description
`stepExecution`	40%	Share of playbook-defined steps that actually ran (0–100)
`stepQuality`	20%	Quality score reflecting step outcomes and whether the configured engines were called (0–100)
`outputCompleteness`	40%	Share of output schema fields that were populated (0–100)

score = round(stepExecution × 0.4 + stepQuality × 0.2 + outputCompleteness × 0.4)

For example, components of 100 / 97.22 / 76.98 give a score of 90 and a level of HIGH.

If any input is null because it could not be measured, the score is normalized over the weights of the remaining inputs. When all three are null, score and level are null as well.

`details[]` — what lowered the score

Individual items that reduced the confidence score. Empty array when nothing was deducted.

Field	Type	Description
`category`	`string`	Reason class: `STEP` / `JSON_OUTPUT` / `TOOL`
`type`	`string`	Specific type (see table below)
`target`	`string`	Subject of the item — an action name or an output schema field path

`type`	`category`	Meaning
`STEP_NEED_REVIEW`	`STEP`	The step finished needing manual review
`STEP_FAILED`	`STEP`	The step finished as failed
`STEP_NOT_EXECUTED`	`STEP`	Defined in the playbook but never executed
`TOOL_NOT_EXECUTED`	`TOOL`	Configured on the workflow but never called
`OUTPUT_MISSING`	`JSON_OUTPUT`	No value was produced for an output schema field
`OUTPUT_EMPTY`	`JSON_OUTPUT`	The field exists but its value is empty

`riskAssessment`

Risk level derived from the overall analysis result.

Field	Type	Description
`riskLevel`	`string`	Risk band derived from `riskScore`: `low` (0–25) / `medium` (26–50) / `high` (51–75) / `critical` (76–100)
`riskScore`	`number`	Risk score (0–100, higher = riskier)
`riskFactors`	`string[]`	Factors that raised the score. Empty array when none

`systemMetadata` — execution metadata

Internal execution information captured while the AI agent ran the analysis.

Field	Type	Description
`totalIterations`	`number`	Total agent iterations. Not the same as step count
`completedWorkIds`	`number[]`	IDs of completed work units
`workflowHistory`	`object[]`	Execution summary per work unit
`tokenUsage`	`object`	Token usage for the whole analysis
`tokenUsagePerWorkId`	`object`	Map of work ID to that unit’s token usage

`systemMetadata.workflowHistory[]`

Only executed work units appear here. Steps defined in the playbook but never executed are absent. To get the full step definition, also fetch GET /workflows/:workflowId and read workflowActions.

Field	Type	Description
`workId`	`number`	Work unit number (starts at 1)
`actionName`	`string`	Executed action function name
`actionDescription`	`string`	Description of the action
`query`	`string`	Query or instruction the agent used for the unit
`reason`	`string`	The agent’s reasoning for the outcome
`success`	`boolean`	Whether the unit succeeded
`verificationStatus`	`string`	Per-action verification result: `passed` / `needs_review` / `failed`
`iteration`	`number`	Iteration number at which the unit completed
`timestamp`	`string`	Completion time (ISO 8601)

`systemMetadata.tokenUsage`

Field	Type	Description
`totalTokens`	`number`	Total tokens
`promptTokens`	`number`	Input (prompt) tokens
`completionTokens`	`number`	Output (completion) tokens

tokenUsagePerWorkId maps a work ID to an object with the same shape.

{
  "tokenUsagePerWorkId": {
    "1": { "totalTokens": 12480, "promptTokens": 11902, "completionTokens": 578 },
    "2": { "totalTokens": 9310, "promptTokens": 9074, "completionTokens": 236 }
  }
}

`outputSchema` · `extractedData` · `extractionStatus` — the output schema area

The internal structure of these three fields follows your workflow’s output schema entirely. There is no common field set defined by Omni. A different workflow means different key names and nesting.

Field	Type	Description
`outputSchema`	`object`	The workflow’s output schema definition at analysis time (JSON Schema), shaped as `{ "type": "object", "properties": { ... } }`
`extractedData`	`object`	Final result populated against that structure. This is the field to consume
`extractionStatus`	`object`	Per top-level schema field extraction state. Values are `extracted` or `missing`

outputSchema is a snapshot frozen at analysis time. Editing the workflow’s output schema afterwards does not change the outputSchema stored on past analyses.

Fields with no value

When the AI cannot find a value for a schema field in the documents, that field is not populated in extractedData and is marked missing in extractionStatus. An OUTPUT_MISSING entry is added to details[] at the same time, lowering the confidence.components.outputCompleteness score.

{
  "extractedData": {
    "corporate_info": {
      "company_name": "Northwind Trading Ltd",
      "registration_number": "00-00-00000"
    }
  },
  "extractionStatus": {
    "corporate_info": "extracted",
    "representative_info": "missing"
  },
  "details": [
    { "category": "JSON_OUTPUT", "type": "OUTPUT_MISSING", "target": "representative_info" }
  ]
}

The keys of extractedData always match the top-level field names of your output schema. Use extractionStatus to determine which fields were actually populated.

`rawActionResults` — raw results per action

Raw output returned by each AI action. Keys are action names (actionName), and which actions exist depends on the workflow’s workflowActions definition.

Field	Type	Description
`answer`	`string`	Action result text
`workId`	`number`	Work unit number
`confidence`	`number`	Result confidence (0.0 – 1.0)
`citationsCount`	`number`	Number of citation chunks used to build the answer
`verificationStatus`	`string`	Action verification result: `passed` / `needs_review` / `failed`

`agentAuditLog` — agent execution audit log

Detailed execution log of every step the AI agent performed. More granular than rawActionResults and includes external engine calls.

agentAuditLog is an object, not an array, and its inner fields use snake_case (executed_at, item_ids, duration_ms). This differs from the camelCase convention at the response top level — take care when parsing.

{
  "agentAuditLog": {
    "steps": [ /* per-step log */ ],
    "summary": { /* aggregate */ }
  }
}

`agentAuditLog.steps[]`

Field	Type	Description
`step`	`number`	Step number (starts at 1)
`action`	`string`	Executed action name
`rule`	`string`	Description of the verification rule the step carried out
`status`	`string`	Step outcome: `passed` / `needs_review` / `failed`
`category`	`string`	Step class: `DOCUMENT` / `IDENTITY` / `COMPLIANCE` / `FINANCIAL` / `OTHER`. May be an empty string for unclassified steps
`tokens`	`number`	Tokens consumed by the step
`item_ids`	`string[]`	IDs of items the step referenced
`mcpcalls`	`object[]`	External engine calls. Empty array when none
`reasoning`	`string`	The agent’s reasoning
`executed_at`	`string`	Execution time (ISO 8601)

`agentAuditLog.steps[].mcpcalls[]`

Populated only for steps that called an external engine (AML screening, text comparison, and so on).

Field	Type	Description
`tool`	`string`	Tool identifier that was called
`engine`	`string`	Engine display name (for example `Text Verifier`, `AML - Business`)
`input`	`object`	Input passed to the tool
`result`	`object`	Raw response from the tool. The shape differs per engine
`purpose`	`string`	Why the agent called this tool
`duration_ms`	`number`	Call duration in milliseconds

`agentAuditLog.summary`

Field	Type	Description
`total_steps`	`number`	Total steps executed
`passed`	`number`	Steps that ended `passed`
`needs_review`	`number`	Steps that ended `needs_review`
`failed`	`number`	Steps that ended `failed`
`total_tokens`	`number`	Token usage for the whole analysis
`overall_decision`	`string`	Aggregate verdict over step outcomes: `passed` / `needs_review` / `failed`
`started_at`	`string`	Start time of the first step (ISO 8601)
`ended_at`	`string`	End time of the last step (ISO 8601)
`subject`	`object`	Summary information about the analysis subject

`findings[]` — verification result items

An ordered summary of each action’s result, intended for display.

Field	Type	Description
`id`	`string`	Unique finding ID (`af_` prefix)
`category`	`string`	Category (same as the action name)
`result`	`string`	Result: `passed` / `warning` / `failed`
`confidence`	`number`	Confidence (0.0 – 1.0)
`details`	`string`	Result details (may be truncated)
`sortOrder`	`number`	Display order

result corresponds to agentAuditLog.steps[].status — passed → passed, needs_review → warning, failed → failed.

`recommendations[]`

Follow-up actions generated by the system.

Field	Type	Description
`id`	`string`	Unique recommendation ID (`ar_` prefix)
`content`	`string`	Recommendation text
`priority`	`number`	Priority (lower means higher priority)
`sortOrder`	`number`	Display order

`targetItems[]` — analyzed items

Items referenced by this analysis.

Field	Type	Description
`itemId`	`string`	Unique item ID (`item_` prefix)
`name`	`string`	Item name
`type`	`string`	Item type: `file` / `text` / `json`
`sourceRef`	`string \| null`	Source reference

Example response

The example below is fictional and exists to illustrate the structure. The fields inside outputSchema and extractedData differ per workflow.

{
  "id": "analysis_8k2m4p9x1q3z",
  "profileId": "pf_3n7v2c5w8k1d",
  "folderId": null,
  "engineId": null,
  "playbookId": "PB-20260701-AB12CD",
  "engine": null,
  "status": "completed",
  "processingTimeMs": 94210,
  "verificationStatus": "pending_review",
  "confidenceScore": 0.6,
  "confidence": {
    "score": 78,
    "level": "HIGH",
    "components": {
      "stepExecution": 100,
      "stepQuality": 87.5,
      "outputCompleteness": 60
    }
  },
  "details": [
    { "category": "STEP", "type": "STEP_NEED_REVIEW", "target": "verify_aml_business" },
    { "category": "JSON_OUTPUT", "type": "OUTPUT_MISSING", "target": "corporate_info.incorporation_date" }
  ],
  "riskAssessment": {
    "riskLevel": "medium",
    "riskScore": 45,
    "riskFactors": ["1 action(s) need manual review"]
  },
  "systemMetadata": {
    "totalIterations": 9,
    "completedWorkIds": [1, 2, 3],
    "workflowHistory": [
      {
        "workId": 1,
        "actionName": "verify_document_presence",
        "actionDescription": "Check that all required documents were submitted",
        "query": "Confirm every required document is present",
        "reason": "All three required documents were confirmed.",
        "success": true,
        "verificationStatus": "passed",
        "iteration": 1,
        "timestamp": "2026-07-02T04:11:23.180Z"
      }
    ],
    "tokenUsage": {
      "totalTokens": 88420,
      "promptTokens": 86104,
      "completionTokens": 2316
    },
    "tokenUsagePerWorkId": {
      "1": { "totalTokens": 12480, "promptTokens": 11902, "completionTokens": 578 }
    }
  },
  "outputSchema": {
    "type": "object",
    "properties": {
      "corporate_info": {
        "type": "object",
        "description": "Corporate information extracted from the submitted documents",
        "properties": {
          "company_name": { "type": "string", "description": "Legal entity name" },
          "registration_number": { "type": "string", "description": "Registration number" }
        }
      }
    }
  },
  "extractedData": {
    "corporate_info": {
      "company_name": "Northwind Trading Ltd",
      "registration_number": "00-00-00000"
    }
  },
  "extractionStatus": {
    "corporate_info": "extracted"
  },
  "rawActionResults": {
    "verify_document_presence": {
      "answer": "All three required documents were submitted.",
      "workId": 1,
      "confidence": 0.9,
      "citationsCount": 0,
      "verificationStatus": "passed"
    }
  },
  "agentAuditLog": {
    "steps": [
      {
        "step": 1,
        "action": "verify_document_presence",
        "rule": "Check that all required documents were submitted",
        "status": "passed",
        "category": "DOCUMENT",
        "tokens": 12480,
        "item_ids": ["item_9d3f6g8h2j4k"],
        "mcpcalls": [
          {
            "tool": "text-similarity",
            "engine": "Text Verifier",
            "input": { "text1": "Business registration certificate", "text2": "business_registration.pdf" },
            "result": { "score": 87 },
            "purpose": "Match submitted file names against required document names",
            "duration_ms": 1840
          }
        ],
        "reasoning": "All three required documents were confirmed.",
        "executed_at": "2026-07-02T04:11:23.180Z"
      }
    ],
    "summary": {
      "total_steps": 3,
      "passed": 2,
      "needs_review": 1,
      "failed": 0,
      "total_tokens": 88420,
      "overall_decision": "needs_review",
      "started_at": "2026-07-02T04:11:23.180Z",
      "ended_at": "2026-07-02T04:12:57.402Z",
      "subject": {}
    }
  },
  "findings": [
    {
      "id": "af_5q1w7e3r9t2y",
      "category": "verify_document_presence",
      "result": "passed",
      "confidence": 0.9,
      "details": "All three required documents were submitted.",
      "sortOrder": 1
    }
  ],
  "recommendations": [
    {
      "id": "ar_4z8x2c6v0b5n",
      "content": "Manual review required - automated checks were inconclusive.",
      "priority": 1,
      "sortOrder": 1
    }
  ],
  "error": null,
  "options": null,
  "clientMetadata": null,
  "targetItems": [
    {
      "itemId": "item_9d3f6g8h2j4k",
      "name": "business_registration.pdf",
      "type": "file",
      "sourceRef": null
    }
  ],
  "primaryReportId": "rpt_2q7w4e9r1t5y",
  "reportId": "rpt_2q7w4e9r1t5y",
  "requestedAt": "2026-07-02T04:10:58.004Z",
  "completedAt": "2026-07-02T04:13:12.661Z",
  "createdAt": "2026-07-02T04:10:58.004Z"
}

Enum reference

Analysis status
Verification verdict
Step status
Finding result
Risk level

Value	Description
`pending`	Waiting to start
`processing`	In progress
`completed`	Finished
`failed`	Failed

System verdict for the analysis as a whole.

Value	Description
`verified`	Every step passed
`pending_review`	At least one step needs manual review
`rejected`	At least one step failed

Execution result of a single playbook action. agentAuditLog.summary.overall_decision, which aggregates them, uses the same value set.

Value	Description
`passed`	Verification passed
`needs_review`	Manual review needed
`failed`	Verification failed

Value	Description
`passed`	Normal
`warning`	Warning (review needed)
`failed`	Failed

Value	Description
`low`	Low risk (0–25)
`medium`	Medium risk (26–50)
`high`	High risk (51–75)
`critical`	Critical (76–100)

Overview

API Reference

Endpoint

Request

Response structure — two distinct groups

System fields

Output schema fields

System fields

Output schema fields

Top-level fields

Three kinds of verification status

`confidenceScore` vs `confidence` — different metrics

`confidence` — Analysis Score

`confidence.components`

`details[]` — what lowered the score

`riskAssessment`

`systemMetadata` — execution metadata

`systemMetadata.workflowHistory[]`

`systemMetadata.tokenUsage`

`outputSchema` · `extractedData` · `extractionStatus` — the output schema area

Fields with no value

`rawActionResults` — raw results per action

`agentAuditLog` — agent execution audit log

`agentAuditLog.steps[]`

`agentAuditLog.steps[].mcpcalls[]`

`agentAuditLog.summary`

`findings[]` — verification result items

`recommendations[]`

`targetItems[]` — analyzed items

Example response

Enum reference

​Endpoint

​Request

​Response structure — two distinct groups

System fields

Output schema fields

​System fields

​Output schema fields

​Top-level fields

​Three kinds of verification status

​confidenceScore vs confidence — different metrics

​confidence — Analysis Score

​confidence.components

​details[] — what lowered the score

​riskAssessment

​systemMetadata — execution metadata

​systemMetadata.workflowHistory[]

​systemMetadata.tokenUsage

​outputSchema · extractedData · extractionStatus — the output schema area

​Fields with no value

​rawActionResults — raw results per action

​agentAuditLog — agent execution audit log

​agentAuditLog.steps[]

​agentAuditLog.steps[].mcpcalls[]

​agentAuditLog.summary

​findings[] — verification result items

​recommendations[]

​targetItems[] — analyzed items

​Example response

​Enum reference

Endpoint

Request

Response structure — two distinct groups

System fields

Output schema fields

Top-level fields

Three kinds of verification status

`confidenceScore` vs `confidence` — different metrics

`confidence` — Analysis Score

`confidence.components`

`details[]` — what lowered the score

`riskAssessment`

`systemMetadata` — execution metadata

`systemMetadata.workflowHistory[]`

`systemMetadata.tokenUsage`

`outputSchema` · `extractedData` · `extractionStatus` — the output schema area

Fields with no value

`rawActionResults` — raw results per action

`agentAuditLog` — agent execution audit log

`agentAuditLog.steps[]`

`agentAuditLog.steps[].mcpcalls[]`

`agentAuditLog.summary`

`findings[]` — verification result items

`recommendations[]`

`targetItems[]` — analyzed items

Example response

Enum reference