@grepr/cli 1.11.6-eb60d62 → 1.11.7-5712be3

package/build/dist/openapi-docs/schemas/Operation.md

@@ -36,237 +36,222 @@ This schema accepts one of the following types:
 `MaskingOperator`
 
 ### Option 10
-`LlmPrompt`
-
-### Option 11
 `PatternMatcher`
 
-### Option 12
+### Option 11
 `LogsIcebergTableSink`
 
-### Option 13
+### Option 12
 `PatternLookupIcebergTableSink`
 
-### Option 14
+### Option 13
 `EventDedupIcebergTableSink`
 
-### Option 15
+### Option 14
 `VendorLogEventDedupIcebergTableSink`
 
-### Option 16
+### Option 15
 `SpanDedupIcebergTableSink`
 
-### Option 17
+### Option 16
 `SpansIcebergTableSink`
 
-### Option 18
+### Option 17
 `LogsIcebergTableSource`
 
-### Option 19
+### Option 18
 `AgentSessionsIcebergTableSource`
 
-### Option 20
+### Option 19
 `AgentSessionEventsIcebergTableSource`
 
-### Option 21
-`LlmPromptResultsIcebergTableSource`
-
-### Option 22
+### Option 20
 `MetricsIcebergTableSource`
 
-### Option 23
+### Option 21
 `TracesIcebergTableSource`
 
-### Option 24
+### Option 22
 `SpansBackfillIcebergTableSource`
 
-### Option 25
+### Option 23
 `MetricsIcebergTableSink`
 
-### Option 26
+### Option 24
 `MetricDataRawIcebergTableSink`
 
-### Option 27
+### Option 25
 `DatadogLogCloudSource`
 
-### Option 28
+### Option 26
 `DatadogLogAgentSource`
 
-### Option 29
+### Option 27
 `DatadogTraceAgentSource`
 
-### Option 30
+### Option 28
 `GreprVendorSource`
 
-### Option 31
+### Option 29
 `GreprMetricsSource`
 
-### Option 32
+### Option 30
 `DatadogLogSink`
 
-### Option 33
+### Option 31
 `DatadogTraceSink`
 
-### Option 34
+### Option 32
 `NewRelicLogSink`
 
-### Option 35
+### Option 33
 `SplunkLogSink`
 
-### Option 36
+### Option 34
 `SumoLogSource`
 
-### Option 37
+### Option 35
 `SumoLogSink`
 
-### Option 38
+### Option 36
 `OtlpLogSink`
 
-### Option 39
+### Option 37
 `OtlpTraceSink`
 
-### Option 40
+### Option 38
 `IcebergLogsReplaySource`
 
-### Option 41
+### Option 39
 `LogsBackfillFlinkSource`
 
-### Option 42
+### Option 40
 `LogsRuleEngine`
 
-### Option 43
+### Option 41
 `ReducerLogsQuerySource`
 
-### Option 44
+### Option 42
 `LogsSynchronousSink`
 
-### Option 45
+### Option 43
 `AgentSessionsSynchronousSink`
 
-### Option 46
+### Option 44
 `MetricsSynchronousSink`
 
-### Option 47
+### Option 45
 `VariantSynchronousSink`
 
-### Option 48
+### Option 46
 `SpansSynchronousSink`
 
-### Option 49
+### Option 47
 `QuerySink`
 
-### Option 50
+### Option 48
 `AgentSignalSink`
 
-### Option 51
+### Option 49
 `DiscardingSink`
 
-### Option 52
+### Option 50
 `DatadogMetricsSink`
 
-### Option 53
+### Option 51
 `LegacyDatadogMetricsSink`
 
-### Option 54
+### Option 52
 `AthenaAnalyticsSource`
 
-### Option 55
+### Option 53
 `GreprReducerLogSource`
 
-### Option 56
+### Option 54
 `AthenaAgentSessionsSource`
 
-### Option 57
+### Option 55
 `GreprRawLogsSource`
 
-### Option 58
+### Option 56
 `GreprRawSpanSource`
 
-### Option 59
-`GreprLlmPromptResultsSource`
-
-### Option 60
+### Option 57
 `TrinoAnalyticsSource`
 
-### Option 61
+### Option 58
 `TrinoReducerLogSource`
 
-### Option 62
+### Option 59
 `TrinoRawLogsSource`
 
-### Option 63
+### Option 60
 `TrinoRawSpanSource`
 
-### Option 64
-`TrinoLlmPromptResultsSource`
-
-### Option 65
+### Option 61
 `SplunkLogAgentSource`
 
-### Option 66
+### Option 62
 `SplunkLogHttpSource`
 
-### Option 67
+### Option 63
 `NewRelicLogAgentSource`
 
-### Option 68
+### Option 64
 `OtlpLogAgentSource`
 
-### Option 69
+### Option 65
 `OtlpTraceAgentSource`
 
-### Option 70
+### Option 66
 `TraceSampler`
 
-### Option 71
+### Option 67
 `SpanTagger`
 
-### Option 72
+### Option 68
 `S3LogsFileSource`
 
-### Option 73
+### Option 69
 `CloudTrailLogsFileSource`
 
-### Option 74
+### Option 70
 `GreprUploadedLogFileSource`
 
-### Option 75
+### Option 71
 `TemplateOperation`
 
-### Option 76
+### Option 72
 `SqlOperation`
 
-### Option 77
+### Option 73
 `TraceReducer`
 
-### Option 78
+### Option 74
 `MetricReducer`
 
-### Option 79
+### Option 75
 `DatadogStatsSink`
 
-### Option 80
+### Option 76
 `LogsValuesSource`
 
-### Option 81
+### Option 77
 `LogRulesApplication`
 
-### Option 82
+### Option 78
 `TriggerActionOp`
 
-### Option 83
-`EntityContextAggregation`
-
-### Option 84
+### Option 79
 `MultiGrokParser`
 
-### Option 85
+### Option 80
 `AgentSessionAnalytics`
 
-### Option 86
+### Option 81
 `AgentLogProcessor`
 
-### Option 87
+### Option 82
 `AgentSpanProcessor`
 
 ## Properties

package/build/dist/openapi-docs/schemas/PatternMatcher.md

@@ -4,6 +4,10 @@
 
 ## Properties
 
+- name: attributeTokenization (optional)
+  type: enum: `KEYS_ONLY`, `KEYS_AND_VALUES`
+  description: Use attribute shape only, or shape and whole-value similarity tokens.
+
 - name: delimiters (required)
   type: array of string
   description: Characters to use for splitting tokens.
@@ -16,6 +20,14 @@
   type: array of string
   description: Masks that are enabled for tokenizing messages.
 
+- name: excludedAttributePrefixes (optional)
+  type: array of array of string
+  description: Excluded attribute prefixes as arrays of literal key segments. Does not affect partition-by values or output attributes.
+
+- name: includedAttributePrefixes (optional)
+  type: array of array of string
+  description: Included attribute prefixes as arrays of literal key segments. Empty selects all. Array elements are traversed without indices.
+
 - name: masks (required)
   type: array of array of string
   description: Set of masks that can be applied to messages before tokenizing them. These masks are specified as pairs of (name, regex).
@@ -37,6 +49,10 @@
   type: number
   description: Controls how close a set of tokens are to a pattern before it is considered a match. Calculated as the percent of matching tokens between a pattern and a message.
 
+- name: tokenizationInput (optional)
+  type: enum: `MESSAGE_ONLY`, `ATTRIBUTES_ONLY`, `MESSAGE_AND_ATTRIBUTES`, `AUTO`
+  description: Selects message text, attributes, both, or message text with an attributes fallback.
+
 - name: type (required)
   type: enum: `pattern-matcher`
   description: Matches log patterns using an iterative feedback loop. Events with known patterns skip the global reducer for improved efficiency.

package/build/dist/openapi-docs/schemas/PromotionPolicy.md

@@ -7,6 +7,10 @@
 - name: attributes (required)
   type: `PromotionPolicyAttributes`
 
+- name: columnRemovalDelay (optional)
+  type: string
+  description: How long after a path or key is excluded its index columns are removed from the table, as an ISO-8601 duration. Defaults to 6 hours when absent.
+
 - name: revision (optional)
   type: integer
   description: Which store of the table's policy this is, assigned by the server: 0 when the table has none. Echo it on a PUT to be refused with 409 if the policy changed since it was read; omit it to replace whatever is stored.

package/build/dist/openapi-docs/schemas/PromotionPolicyAttributeExclude.md

@@ -1,5 +1,7 @@
 # PromotionPolicyAttributeExclude
 
+An attribute path whose columns writers stop filling, unless an include covers it, and which are removed from the table at scheduledRemovalAt. No exclude entry is allowed while excludeAllByDefault is on.
+
 **Type**: `object`
 
 ## Properties
@@ -9,3 +11,7 @@
 
 - name: prefix (optional)
   type: boolean
+
+- name: scheduledRemovalAt (optional)
+  type: string
+  description: When the columns this entry excludes are removed from the table, assigned by the server. Taking the entry off the exclude list before then keeps them.

package/build/dist/openapi-docs/schemas/PromotionPolicyAttributeInclude.md

@@ -1,11 +1,14 @@
 # PromotionPolicyAttributeInclude
 
+An attribute path indexed whatever an exclude entry or excludeAllByDefault says. It counts against the cap: an exact entry gets its columns when the policy is stored, if the cap has room, and the paths beneath a nested entry are discovered like any other.
+
 **Type**: `object`
 
 ## Properties
 
 - name: numeric (optional)
   type: boolean
+  description: Whether the path gets a numeric index column when the policy is stored, without waiting for a number. Exact entries only: refused when prefix is true.
 
 - name: path (required)
   type: string

package/build/dist/openapi-docs/schemas/PromotionPolicyAttributes.md

@@ -10,5 +10,13 @@
 - name: exclude (optional)
   type: array of `PromotionPolicyAttributeExclude`
 
+- name: excludeAllByDefault (optional)
+  type: boolean
+  description: Whether every path no include covers is excluded. Writers stop filling the existing columns no include covers, and they are removed from the table at excludeAllScheduledRemovalAt.
+
+- name: excludeAllScheduledRemovalAt (optional)
+  type: string
+  description: When the columns excludeAllByDefault excludes are removed from the table, assigned by the server. Turning excludeAllByDefault off before then keeps them.
+
 - name: include (optional)
   type: array of `PromotionPolicyAttributeInclude`

package/build/dist/openapi-docs/schemas/PromotionPolicyTags.md

@@ -12,3 +12,7 @@
 
 - name: include (optional)
   type: array of string
+
+- name: scheduledRemovalAt (optional)
+  type: object
+  description: When each excluded key's column is removed from the table, assigned by the server. Taking the key off the exclude list before then keeps the column.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@grepr/cli",
-  "version": "1.11.6-eb60d62",
+  "version": "1.11.7-5712be3",
   "description": "Official Grepr CLI tool for submitting synchronous batch jobs with OAuth authentication",
   "main": "build/dist/grepr.js",
   "type": "module",

package/build/dist/docs-index/00b1368c-5d7a-44cc-b31d-a1bb14e0a94c.txt

@@ -1,43 +0,0 @@
-# LlmPromptResultsIcebergTableSource
-
-**Type**: `object`
-
-## Properties
-
-- name: datasetId (required)
-  type: string
-  description: The ID of the dataset to read data from.
-
-- name: end (required)
-  type: string
-  description: End of time interval to query data for.
-
-- name: limit (optional)
-  type: integer
-  description: The maximum number of rows to process
-
-- name: name (required)
-  type: string
-
-- name: offset (optional)
-  type: integer
-  description: Offset on the number of rows to process
-
-- name: query (required)
-  type: `query`
-
-- name: sortOrder (optional)
-  type: enum: `ASCENDING`, `DESCENDING`, `UNSORTED`
-  description: The order in which the rows should be sorted by
-
-- name: start (required)
-  type: string
-  description: Start of time interval to query data for.
-
-- name: type (required)
-  type: enum: `llm-prompt-results-iceberg-table-source`
-  description: Reads LLM prompt results from an Iceberg table for processing.
-
-- name: variables (optional)
-  type: object
-  description: Variables that can be used to modify the query while parsed

package/build/dist/docs-index/068baa5d-1f8e-4e2b-a522-9322810bb1e0.txt

@@ -1,11 +0,0 @@
-# Simple
-
-Simple workflow that sends the log event directly to the LLM.
-
-**Type**: `object`
-
-## Properties
-
-- name: type (required)
-  type: enum: `simple`
-  description: Simple workflow that sends the log event directly to the LLM.

package/build/dist/docs-index/16dae2a3-9ab0-44e6-8564-469bb2e96192.txt

@@ -1,61 +0,0 @@
-# LlmPrompt
-
-**Type**: `object`
-
-## Properties
-
-- name: availableRuleTypes (optional)
-  type: array of string
-  description: Rule types that the LLM can produce as output. When non-empty, the LLM output schema is augmented with a 'rules' field describing the available rule type schemas. Rules returned by the LLM are parsed and persisted to the LogRulesApplication operator for dynamic per-pattern rule application.
-
-- name: availableTools (optional)
-  type: array of `LlmToolConfig`
-  description: MCP tool integrations available for the LLM to invoke as terminal actions. Each entry references an integration that exposes MCP-compatible tools.
-
-- name: capacity (optional)
-  type: integer
-  description: Maximum number of concurrent LLM calls per operator instance. Note: Total concurrent calls = capacity × Flink parallelism. For example, with parallelism=4 and capacity=100, up to 400 concurrent calls may occur. Consider your LLM provider's rate limits when configuring this value.
-
-- name: integrationId (required)
-  type: string
-  description: The LLM integration ID (Gemini or OpenAI).
-
-- name: llmConfig (required)
-  type: `LlmConfig`
-
-- name: maxRetries (optional)
-  type: integer
-  description: Maximum number of retries for failed LLM calls.
-
-- name: name (required)
-  type: string
-
-- name: outputActions (optional)
-  type: array of enum: `FORWARD_TO_SINKS`
-  description: Controls where LLM-enriched events are routed after processing. Multiple actions can be combined.
-
-- name: outputSchema (optional)
-  type: string
-  description: JSON Schema defining the expected structure of the LLM output. When set, the LLM is instructed to return structured JSON matching this schema. The output will include an 'actionable' boolean and a 'user_result' object. Supports standard JSON Schema property keywords: type, description, and enum.
-
-- name: predicate (optional)
-  type: `EventPredicate`
-
-- name: promptInstruction (required)
-  type: string
-  description: The user's prompt instruction describing what the LLM should do. Context (event JSON, entity context, similar patterns) is assembled automatically by the system — do not include placeholders.
-
-- name: responseFormat (optional)
-  type: enum: `TEXT`, `JSON`
-  description: The format for the LLM response.
-
-- name: timeoutSeconds (optional)
-  type: integer
-  description: Timeout for LLM calls in seconds.
-
-- name: type (required)
-  type: enum: `llm-prompt`
-  description: Runs LLM prompts on new log patterns.
-
-- name: workflowConfig (optional)
-  type: `LlmWorkflowConfig`

package/build/dist/docs-index/609a0550-6cff-4d28-acf1-453c1a6f14ab.txt

@@ -1,11 +0,0 @@
-# PromotionPolicyAttributeExclude
-
-**Type**: `object`
-
-## Properties
-
-- name: path (required)
-  type: string
-
-- name: prefix (optional)
-  type: boolean

package/build/dist/docs-index/654d5da3-7e4a-47ad-b76c-50ba9312eea1.txt

@@ -1,15 +0,0 @@
-# TriggerActionConfig
-
-Configuration for the trigger action operator in an AI pipeline.
-
-**Type**: `object`
-
-## Properties
-
-- name: capacity (optional)
-  type: integer
-  description: Maximum number of concurrent trigger handler calls.
-
-- name: timeoutSeconds (optional)
-  type: integer
-  description: Timeout for trigger handler execution in seconds.

package/build/dist/docs-index/8706eefa-5f8f-42a3-9269-9ae9a1d833f5.txt

@@ -1,15 +0,0 @@
-# LlmWorkflowConfig
-
-Configuration for the LLM workflow strategy.
-
-**Type**: `object`
-
-## One Of
-
-This schema accepts one of the following types:
-
-### Option 1
-`Simple`
-
-### Option 2
-`Rag`

package/build/dist/docs-index/92ab8f13-d673-4c48-8e3f-fa95073781f4.txt

@@ -1,45 +0,0 @@
-# EntityContextAggregation
-
-**Type**: `object`
-
-## Properties
-
-- name: capacity (optional)
-  type: integer
-  description: Maximum number of concurrent LLM calls per operator instance.
-
-- name: entityAttributes (optional)
-  type: array of string
-
-- name: entityTags (optional)
-  type: array of string
-
-- name: integrationId (required)
-  type: string
-  description: The LLM integration ID (Gemini or OpenAI).
-
-- name: llmConfig (required)
-  type: `LlmConfig`
-
-- name: maxRetries (optional)
-  type: integer
-  description: Maximum retries for failed LLM calls.
-
-- name: maxSummaryTokens (optional)
-  type: integer
-  description: Maximum token budget for the entity summary. The system prompt instructs the LLM to keep the summary within this bound.
-
-- name: name (required)
-  type: string
-
-- name: systemPromptOverride (optional)
-  type: string
-  description: Optional system prompt override. When null, a default system prompt is used that instructs the LLM to maintain a rolling entity summary and report whether the update is broadcast-worthy.
-
-- name: timeoutSeconds (optional)
-  type: integer
-  description: Timeout for LLM calls in seconds.
-
-- name: type (required)
-  type: enum: `entity-context-aggregation`
-  description: Maintains per-entity rolling context summaries using an LLM. Broadcasts significant updates for downstream prompt enrichment.

package/build/dist/docs-index/946ebf1b-3907-4fe2-930c-5028551b5741.txt

@@ -1,43 +0,0 @@
-# GreprLlmPromptResultsSource
-
-**Type**: `object`
-
-## Properties
-
-- name: datasetId (required)
-  type: string
-  description: The ID of the dataset to read data from.
-
-- name: end (required)
-  type: string
-  description: End of time interval to query data for.
-
-- name: limit (optional)
-  type: integer
-  description: The maximum number of rows to process
-
-- name: name (required)
-  type: string
-
-- name: offset (optional)
-  type: integer
-  description: Offset on the number of rows to process
-
-- name: query (required)
-  type: `query`
-
-- name: sortOrder (optional)
-  type: enum: `ASCENDING`, `DESCENDING`, `UNSORTED`
-  description: The order in which the rows should be sorted by
-
-- name: start (required)
-  type: string
-  description: Start of time interval to query data for.
-
-- name: type (required)
-  type: enum: `grepr-llm-prompt-results-source`
-  description: Reads LLM prompt result events using Grepr's internal query system from an Iceberg Table.
-
-- name: variables (optional)
-  type: object
-  description: Variables that can be used to modify the query while parsed

package/build/dist/docs-index/9ae737d1-c03b-407b-8df7-9a61fdc6be65.txt

@@ -1,16 +0,0 @@
-# EmbeddingConfig
-
-Configuration for an embedding model. The 'type' field determines the provider.
-
-**Type**: `object`
-
-## Properties
-
-- name: dimensions (optional)
-  type: integer
-
-- name: model (optional)
-  type: string
-
-- name: type (required)
-  type: string