@jambonz/schema 0.3.11 → 0.3.13

package/components/recognizer-sonioxOptions.schema.json

@@ -11,19 +11,91 @@
     },
     "model": {
       "type": "string",
-      "description": "Soniox recognition model."
+      "description": "Soniox recognition model (e.g. 'stt-rt-v5')."
+    },
+    "languageHints": {
+      "type": "array",
+      "items": { "type": "string" },
+      "description": "Language hints for the v5 multilingual model, as ISO language codes (e.g. ['en','es']). Improves accuracy and speeds up language detection."
+    },
+    "enableLanguageIdentification": {
+      "type": "boolean",
+      "description": "Enable language identification (v5); each recognized token is tagged with its detected language."
+    },
+    "enableSpeakerDiarization": {
+      "type": "boolean",
+      "description": "Enable speaker diarization (v5); each recognized token is tagged with a speaker."
     },
     "endpointDetection": {
       "type": "boolean",
       "description": "Enable endpoint detection."
     },
+    "endpointSensitivity": {
+      "type": "number",
+      "minimum": -1,
+      "maximum": 1,
+      "description": "Endpoint detection sensitivity (v5), -1.0 to 1.0 (default 0.0). Higher values finalize endpoints faster; lower values are more conservative."
+    },
+    "maxEndpointDelayMs": {
+      "type": "integer",
+      "minimum": 500,
+      "maximum": 3000,
+      "description": "Maximum delay in milliseconds before an endpoint is forced (v5), 500-3000 (default 2000)."
+    },
+    "maxNonFinalTokensDurationMs": {
+      "type": "integer",
+      "minimum": 0,
+      "description": "Maximum duration in milliseconds that tokens may remain non-final before forced finalization (v5)."
+    },
     "profanityFilter": {
       "type": "boolean",
       "description": "Filter profanity from results."
     },
+    "context": {
+      "type": "object",
+      "additionalProperties": false,
+      "description": "Soniox v5 recognition context to improve accuracy (max ~8000 tokens). See https://soniox.com/docs/stt/concepts/context. When supplied this takes precedence outright over the 'speechContext' and 'hints' shortcuts.",
+      "properties": {
+        "general": {
+          "type": "array",
+          "description": "Structured key/value metadata describing the domain, topic, participants, setting, etc.",
+          "items": {
+            "type": "object",
+            "additionalProperties": false,
+            "properties": {
+              "key": { "type": "string" },
+              "value": { "type": "string" }
+            },
+            "required": ["key", "value"]
+          }
+        },
+        "text": {
+          "type": "string",
+          "description": "Free-form background text (history of prior interactions, reference documents, meeting notes, summaries)."
+        },
+        "terms": {
+          "type": "array",
+          "items": { "type": "string" },
+          "description": "Domain-specific vocabulary to boost recognition."
+        },
+        "translation_terms": {
+          "type": "array",
+          "description": "Source/target term mappings; only used when translation is enabled.",
+          "items": {
+            "type": "object",
+            "additionalProperties": false,
+            "properties": {
+              "source": { "type": "string" },
+              "target": { "type": "string" }
+            },
+            "required": ["source", "target"]
+          }
+        }
+      }
+    },
     "speechContext": {
       "type": "string",
-      "description": "Speech context for improved recognition."
+      "description": "Shortcut for context.text (free-text background). Ignored when the full 'context' object is supplied. The generic recognizer 'hints' are likewise sent as context.terms unless 'context' is supplied."
     },
     "clientRequestReference": {
       "type": "string",

package/components/recognizer.schema.json

@@ -8,7 +8,7 @@
     "vendor": {
       "type": "string",
       "description": "The STT vendor to use. Must match a vendor configured in the jambonz platform.",
-      "examples": ["google", "aws", "microsoft", "deepgram", "nuance", "ibm", "nvidia", "soniox", "cobalt", "assemblyai", "speechmatics", "openai", "houndify", "gladia", "elevenlabs", "verbio", "custom"]
+      "examples": ["google", "aws", "microsoft", "deepgram", "ibm", "nvidia", "soniox", "assemblyai", "speechmatics", "openai", "houndify", "gladia", "elevenlabs", "custom"]
     },
     "label": {
       "type": "string",
@@ -195,18 +195,15 @@
     "googleOptions": { "$ref": "recognizer-googleOptions" },
     "awsOptions": { "$ref": "recognizer-awsOptions" },
     "azureOptions": { "$ref": "recognizer-azureOptions" },
-    "nuanceOptions": { "$ref": "recognizer-nuanceOptions" },
     "ibmOptions": { "$ref": "recognizer-ibmOptions" },
     "nvidiaOptions": { "$ref": "recognizer-nvidiaOptions" },
     "sonioxOptions": { "$ref": "recognizer-sonioxOptions" },
-    "cobaltOptions": { "$ref": "recognizer-cobaltOptions" },
     "assemblyAiOptions": { "$ref": "recognizer-assemblyAiOptions" },
     "speechmaticsOptions": { "$ref": "recognizer-speechmaticsOptions" },
     "openaiOptions": { "$ref": "recognizer-openaiOptions" },
     "houndifyOptions": { "$ref": "recognizer-houndifyOptions" },
     "gladiaOptions": { "$ref": "recognizer-gladiaOptions" },
     "elevenlabsOptions": { "$ref": "recognizer-elevenlabsOptions" },
-    "verbioOptions": { "$ref": "recognizer-verbioOptions" },
     "customOptions": { "$ref": "recognizer-customOptions" }
   },
   "required": ["vendor"],

package/components/synthesizer.schema.json

@@ -8,7 +8,7 @@
     "vendor": {
       "type": "string",
       "description": "The TTS vendor to use. Must match a vendor configured in the jambonz platform.",
-      "examples": ["google", "aws", "microsoft", "elevenlabs", "cartesia", "deepgram", "ibm", "nuance", "nvidia", "wellsaid", "whisper", "verbio", "custom"]
+      "examples": ["google", "aws", "microsoft", "elevenlabs", "cartesia", "deepgram", "ibm", "nvidia", "wellsaid", "whisper", "custom"]
     },
     "label": {
       "type": "string",

package/docs/components/recognizer.md

@@ -65,14 +65,6 @@ See [Voice Agent API documentation](https://docs.speechmatics.com/private/voice-
 
 - [STT Models](https://soniox.com/docs/stt/models)
 
-### Verbio
-
-- [Supported Languages](https://www.verbio.com/supported-languages)
-
 ### Gladia
 
 - [Supported Languages](https://docs.gladia.io/chapters/language/supported-languages)
-
-### Nuance
-
-- [ASR gRPC API](https://docs.mix.nuance.com/asr-grpc/v1/) (Nuance is now part of Microsoft; Azure Speech Service is the successor)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/schema",
-  "version": "0.3.11",
+  "version": "0.3.13",
   "description": "JSON Schema definitions and validation for jambonz verb applications",
   "main": "index.js",
   "scripts": {

package/verbs/agent.schema.json

@@ -199,6 +199,11 @@
               "minimum": 0,
               "description": "Sampling temperature."
             },
+            "reasoningEffort": {
+              "type": "string",
+              "enum": ["minimal", "low", "medium", "high"],
+              "description": "Vendor-neutral thinking/reasoning effort. Mapped per-vendor by the LLM adapter (Gemini thinkingLevel, OpenAI reasoning_effort, Anthropic extended thinking); ignored by vendors without a native equivalent. 'minimal' minimizes thinking for lowest TTFT on latency-sensitive voice turns."
+            },
             "tools": {
               "type": "array",
               "description": "Tool / function definitions available to the model. The MCP-flat shape `{name, description, parameters}` is canonical; the OpenAI-wrapped form `{type:'function', function:{...}}` is also accepted.",

package/components/recognizer-cobaltOptions.schema.json

@@ -1,34 +0,0 @@
-{
-  "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://jambonz.org/schema/components/recognizer-cobaltOptions",
-  "title": "Cobalt Recognizer Options",
-  "description": "Cobalt-specific STT options. Only applies when recognizer vendor is 'cobalt'.",
-  "type": "object",
-  "properties": {
-    "serverUri": {
-      "type": "string",
-      "description": "Cobalt server URI."
-    },
-    "enableConfusionNetwork": {
-      "type": "boolean",
-      "description": "Enable confusion network output."
-    },
-    "metadata": {
-      "type": "string",
-      "description": "Metadata string to pass to the server."
-    },
-    "compiledContextData": {
-      "type": "string",
-      "description": "Compiled context data for biasing recognition."
-    },
-    "wordTimeOffsets": {
-      "type": "boolean",
-      "description": "Include word-level timestamps."
-    },
-    "contextToken": {
-      "type": "string",
-      "description": "Context token for server-side context."
-    }
-  },
-  "additionalProperties": false
-}

package/components/recognizer-nuanceOptions.schema.json

@@ -1,150 +0,0 @@
-{
-  "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://jambonz.org/schema/components/recognizer-nuanceOptions",
-  "title": "Nuance Recognizer Options",
-  "description": "Nuance Mix specific options. Only applies when recognizer vendor is 'nuance'.",
-  "type": "object",
-  "properties": {
-    "clientId": {
-      "type": "string",
-      "description": "Nuance Mix client ID."
-    },
-    "secret": {
-      "type": "string",
-      "description": "Nuance Mix client secret."
-    },
-    "kryptonEndpoint": {
-      "type": "string",
-      "description": "Custom Nuance Krypton endpoint URL."
-    },
-    "topic": {
-      "type": "string",
-      "description": "Recognition topic (domain)."
-    },
-    "utteranceDetectionMode": {
-      "type": "string",
-      "enum": ["single", "multiple", "disabled"],
-      "description": "How utterance boundaries are detected."
-    },
-    "punctuation": {
-      "type": "boolean",
-      "description": "Enable automatic punctuation."
-    },
-    "profanityFilter": {
-      "type": "boolean",
-      "description": "Filter profanity from results."
-    },
-    "includeTokenization": {
-      "type": "boolean",
-      "description": "Include tokenization data in results."
-    },
-    "discardSpeakerAdaptation": {
-      "type": "boolean",
-      "description": "Discard speaker adaptation data."
-    },
-    "suppressCallRecording": {
-      "type": "boolean",
-      "description": "Suppress call recording on the Nuance side."
-    },
-    "maskLoadFailures": {
-      "type": "boolean",
-      "description": "Mask resource load failures."
-    },
-    "suppressInitialCapitalization": {
-      "type": "boolean",
-      "description": "Suppress initial capitalization of results."
-    },
-    "allowZeroBaseLmWeight": {
-      "type": "boolean",
-      "description": "Allow zero base language model weight."
-    },
-    "filterWakeupWord": {
-      "type": "boolean",
-      "description": "Filter wakeup words from results."
-    },
-    "resultType": {
-      "type": "string",
-      "enum": ["final", "partial", "immutable_partial"],
-      "description": "Type of results to return."
-    },
-    "noInputTimeoutMs": {
-      "type": "number",
-      "description": "Timeout in milliseconds before no-input event."
-    },
-    "recognitionTimeoutMs": {
-      "type": "number",
-      "description": "Maximum recognition duration in milliseconds."
-    },
-    "utteranceEndSilenceMs": {
-      "type": "number",
-      "description": "Silence duration in milliseconds to detect end of utterance."
-    },
-    "maxHypotheses": {
-      "type": "number",
-      "description": "Maximum number of recognition hypotheses to return."
-    },
-    "speechDomain": {
-      "type": "string",
-      "description": "Speech domain for optimized recognition."
-    },
-    "formatting": {
-      "type": "object",
-      "description": "Formatting options for recognition results.",
-      "properties": {
-        "scheme": { "type": "string", "description": "Formatting scheme name." },
-        "options": { "type": "object", "description": "Scheme-specific formatting options." }
-      },
-      "required": ["scheme", "options"]
-    },
-    "clientData": {
-      "type": "object",
-      "description": "Custom client data to pass to Nuance.",
-      "additionalProperties": true
-    },
-    "userId": {
-      "type": "string",
-      "description": "User ID for speaker adaptation."
-    },
-    "speechDetectionSensitivity": {
-      "type": "number",
-      "description": "Speech detection sensitivity (0-1)."
-    },
-    "resources": {
-      "type": "array",
-      "description": "Array of Nuance recognition resources (grammars, wordsets, etc.).",
-      "items": {
-        "type": "object",
-        "properties": {
-          "externalReference": {
-            "type": "object",
-            "description": "External resource reference.",
-            "properties": {
-              "type": {
-                "type": "string",
-                "enum": ["undefined_resource_type", "wordset", "compiled_wordset", "domain_lm", "speaker_profile", "grammar", "settings"]
-              },
-              "uri": { "type": "string" },
-              "maxLoadFailures": { "type": "boolean" },
-              "requestTimeoutMs": { "type": "number" },
-              "headers": { "type": "object" }
-            }
-          },
-          "inlineWordset": { "type": "string", "description": "Inline wordset JSON string." },
-          "builtin": { "type": "string", "description": "Built-in grammar name." },
-          "inlineGrammar": { "type": "string", "description": "Inline SRGS grammar." },
-          "wakeupWord": { "type": "array", "items": { "type": "string" }, "description": "Wakeup words." },
-          "weightName": {
-            "type": "string",
-            "enum": ["defaultWeight", "lowest", "low", "medium", "high", "highest"]
-          },
-          "weightValue": { "type": "number" },
-          "reuse": {
-            "type": "string",
-            "enum": ["undefined_reuse", "low_reuse", "high_reuse"]
-          }
-        }
-      }
-    }
-  },
-  "additionalProperties": false
-}

package/components/recognizer-verbioOptions.schema.json

@@ -1,46 +0,0 @@
-{
-  "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://jambonz.org/schema/components/recognizer-verbioOptions",
-  "title": "Verbio Recognizer Options",
-  "description": "Verbio-specific STT options. Only applies when recognizer vendor is 'verbio'.",
-  "type": "object",
-  "properties": {
-    "enable_formatting": {
-      "type": "boolean",
-      "description": "Enable text formatting of results."
-    },
-    "enable_diarization": {
-      "type": "boolean",
-      "description": "Enable speaker diarization."
-    },
-    "topic": {
-      "type": "number",
-      "description": "Topic ID for domain-specific recognition."
-    },
-    "inline_grammar": {
-      "type": "string",
-      "description": "Inline SRGS grammar for constrained recognition."
-    },
-    "grammar_uri": {
-      "type": "string",
-      "description": "URI of an external grammar resource."
-    },
-    "label": {
-      "type": "string",
-      "description": "Label for the recognition session."
-    },
-    "recognition_timeout": {
-      "type": "number",
-      "description": "Maximum recognition duration in seconds."
-    },
-    "speech_complete_timeout": {
-      "type": "number",
-      "description": "Silence duration in seconds after complete speech."
-    },
-    "speech_incomplete_timeout": {
-      "type": "number",
-      "description": "Silence duration in seconds after incomplete speech."
-    }
-  },
-  "additionalProperties": false
-}