audio-transcriber
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| HOST | No | Bind host for HTTP transports | 0.0.0.0 |
| PORT | No | Bind port for HTTP transports | 8000 |
| DEBUG | No | Verbose logging | False |
| MCP_URL | No | URL of the MCP server the agent connects to (agent only) | http://localhost:8000/mcp |
| MISCTOOL | No | Toggle the miscellaneous/health-check tool | True |
| MODEL_ID | No | Model ID for the agent (agent only) | gpt-4o |
| PROVIDER | No | LLM provider for the agent (agent only) | openai |
| TRANSPORT | No | Transport mode: stdio, streamable-http, or sse | stdio |
| ENABLE_OTEL | No | Enable OpenTelemetry export | True |
| EUNOMIA_TYPE | No | Authorization mode: none, embedded, remote | none |
| ENABLE_WEB_UI | No | Serve the AG-UI web interface (agent only) | True |
| MCP_TOOL_MODE | No | Tool surface: condensed, verbose, or both | condensed |
| WHISPER_MODEL | No | Local OpenAI Whisper model for local transcription (e.g., base, tiny, small) | base |
| MCP_ENABLED_TAGS | No | Comma-separated tag allow-list | |
| PYTHONUNBUFFERED | No | Unbuffered stdout | 1 |
| MCP_DISABLED_TAGS | No | Comma-separated tag deny-list | |
| MCP_ENABLED_TOOLS | No | Comma-separated tool allow-list | |
| EUNOMIA_REMOTE_URL | No | Remote Eunomia server URL | |
| MCP_DISABLED_TOOLS | No | Comma-separated tool deny-list | |
| OPENROUTER_API_KEY | Yes | OpenRouter API key for LLM provider | |
| EUNOMIA_POLICY_FILE | No | Embedded policy file | mcp_policies.json |
| AUDIO_PROCESSINGTOOL | No | Toggle the audio-processing (transcription) tool | True |
| AUDIO_TRANSCRIPTOR_API_KEY | Yes | API key for the audio transcription backend | |
| OTEL_EXPORTER_OTLP_ENDPOINT | No | OTLP collector endpoint | |
| OTEL_EXPORTER_OTLP_PROTOCOL | No | OTLP protocol (e.g., http/protobuf) | |
| OTEL_EXPORTER_OTLP_PUBLIC_KEY | No | OTLP auth public key | |
| OTEL_EXPORTER_OTLP_SECRET_KEY | No | OTLP auth secret key | |
| LANGSMITH_DEFAULT_SYSTEM_PROMPT | No | Default system prompt for LangSmith tracing |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"subscribe": false,
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| transcribe_audioC | Transcribes audio from a provided file or by recording from the microphone. |
| health_checkA | Report the transcription engine's state without loading a model. Reads which Whisper backends are importable, whether ffmpeg is on the path, the default model and the transcription directory. Changes nothing. |
| audio_ingest_transcriptionA | Transcribe an audio file and natively ingest it into the knowledge graph. Runs Whisper on |
| transcribe_media_transcribe_segmentsA | Transcribe segments (transcribe_media). Verify the digest, transcribe via Whisper, and return AudioSegment-shaped
segments ( |
| audio_transcriber_exportC | Export transcription to specified formats. |
| audio_transcriber_initiate_streamC | Initiate the audio input stream. |
| audio_transcriber_interactC | Interact with PersonaPlex server via WebSocket. |
| audio_transcriber_recordC | Record audio for a specified duration or until stopped. |
| audio_transcriber_save_streamB | Save the recorded frames to a WAV file. |
| audio_transcriber_stop_streamA | Stop and close the audio stream. |
| audio_transcriber_transcribeC | Transcribe the audio file using the initialized backend. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| connector-status | Ask for this connector server's current operational status. |
| transcription_specialist | Audio transcription specialist: run Whisper over audio/video, export captions, and ingest transcripts into the knowledge graph via the audio-transcriber MCP server. |
| transcribe_file_prompt | Generates a prompt for transcribing a specific audio file. |
| record_and_transcribe_prompt | Generates a prompt for recording from the microphone and transcribing. |
| translate_audio_prompt | Generates a prompt for transcribing an audio file and translating it to English. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| _connector_status_resource | Native resource exposing this server's static status payload. |
| ontology://audio-transcriber-mcp/audio.ttl | |
| connector-resources | Data sources, ontology mappings and resources of this connector. |
| audio-transcriber-operations/SKILL.md | Operate audio-transcriber through its governed MCP and Arbora capabilities, including audio transcriber knowledge graph, audio transcriber transcription. Use when a request requires this provider's read, change, automation, ingestion, troubleshooting, or evidence workflows. |
| audio-transcriber-operations/_manifest | File listing for audio-transcriber-operations |
| audio-transcriber-development/SKILL.md | Develop and maintain the Audio Transcriber connector while preserving its Whisper backend boundary, MCP tool surface, typed transcript ingestion, optional hardware behavior, package resources, and public-first evidence. |
| audio-transcriber-development/_manifest | File listing for audio-transcriber-development |
| audio-transcriber-knowledge-graph/SKILL.md | Native knowledge-graph ingestion of Whisper transcripts on the audio-transcriber MCP server — transcribe an audio/video file and push it into the epistemic-graph in one call: the raw audio as a shared :MediaAsset blob, the transcript text as a :Document, and each Whisper segment as a :TranscriptSegment node linked :segmentOf the transcript and :transcribedFrom the audio. Use when a transcript must be durable and semantically searchable in the KG. Do NOT use for a one-off transcription with no KG persistence (use audio-transcriber-transcription). |
| audio-transcriber-knowledge-graph/_manifest | File listing for audio-transcriber-knowledge-graph |
| audio-transcriber-transcription/SKILL.md | Speech-to-text on the audio-transcriber MCP server — run Whisper (faster-whisper, falling back to openai-whisper) over a local audio/video file or a microphone recording, and export txt/srt/vtt/json captions. Use when the agent must transcribe or translate spoken audio, generate subtitle/caption files, or pick a Whisper model for an accuracy/latency trade-off. Do NOT use to push a transcript into the knowledge graph (use audio-transcriber-knowledge-graph) or to download the media first (use media-downloader). |
| audio-transcriber-transcription/_manifest | File listing for audio-transcriber-transcription |
TDQS
Scored across 11 tools
Several tools perform overlapping transcription: transcribe_audio, audio_transcriber_transcribe, audio_ingest_transcription, and transcribe_media_transcribe_segments all transcribe audio with only subtle boundary differences. Recording is also split between transcribe_audio (microphone) and audio_transcriber_record, making tool selection ambiguous.
The names mix unprefixed verb_noun forms, audio_transcriber_ prefixed names, and awkward compound names like transcribe_media_transcribe_segments. The convention is inconsistent across the set, though most names remain readable.
Eleven tools is well within the appropriate range for an audio transcription server with streaming, export, and ingestion features. The count is not excessive and each tool could plausibly serve a distinct workflow.
The surface covers health checks, file and microphone transcription, streaming controls, export, KG ingestion, and segment output. Minor gaps exist around model configuration, transcript listing, or batch management, but core workflows are represented.