audio-transcriber
README.md
# Audio Transcriber
## CLI or API | MCP | Agent
*Version: 2.1.0*
> **Documentation** — Installation, deployment, and usage across the CLI, Python API,
> MCP server, and A2A agent are maintained in the
> [official documentation](https://knuckles-team.github.io/audio-transcriber/).
---
## Overview
**Audio Transcriber** is a production-grade Agent and Model Context Protocol (MCP) server designed to interface directly with Transcribe your .wav .mp4 .mp3 .flac files to text or record your own audio!.
---
## Key Features
- **Consolidated Action-Routed MCP Tools:** Minimizes token overhead and eliminates tool bloat in LLM contexts by grouping methods into optimized, togglable tool modules.
- **Enterprise-Grade Security:** Comprehensive support for Eunomia policies, OIDC token delegation, and granular execution context tracking.
- **Integrated Graph Agent:** Built-in Pydantic AI agent supporting the Agent Control Protocol (ACP) and standard Web interfaces (AG-UI).
- **Native Telemetry & Tracing:** Out-of-the-box OpenTelemetry exports and native Langfuse tracing.
---
## CLI or API
This agent wraps the Transcribe your .wav .mp4 .mp3 .flac files to text or record your own audio! API. You can interact with it programmatically or via its integrated execution entrypoints.
Detailed instructions on how to use the underlying API wrappers, extended schema bindings, and developer SDK references are maintained in [docs/index.md](docs/index.md).
---
## MCP
This server utilizes dynamic Action-Routed tools to optimize token overhead and maximize IDE compatibility.
### Available MCP Tools
The table below is auto-generated from the live server — do not edit by hand.
<!-- MCP-TOOLS-TABLE:START -->
#### Condensed action-routed tools (`MCP_TOOL_MODE=condensed`)
| MCP Tool | Toggle Env Var | Description |
|----------|----------------|-------------|
| `audio_ingest_transcription` | `MISCTOOL` | Transcribe an audio file and natively ingest it into the knowledge graph. |
| `health_check` | `MISCTOOL` | |
| `transcribe_audio` | `AUDIO_PROCESSINGTOOL` | Transcribes audio from a provided file or by recording from the microphone. |
| `transcribe_media` | `MEDIA_SIDECARTOOL` | Verify the digest, transcribe via Whisper, and return AudioSegment-shaped |
#### Verbose 1:1 API-mapped tools (`MCP_TOOL_MODE=verbose` or `both`)
<details>
<summary>7 per-operation tools — one per public API method (click to expand)</summary>
| MCP Tool | Toggle Env Var | Description |
|----------|----------------|-------------|
| `audio_transcriber_export` | `AUDIO_TRANSCRIBERTOOL` | Export transcription to specified formats. |
| `audio_transcriber_initiate_stream` | `AUDIO_TRANSCRIBERTOOL` | Initiate the audio input stream. |
| `audio_transcriber_interact` | `AUDIO_TRANSCRIBERTOOL` | Interact with PersonaPlex server via WebSocket. |
| `audio_transcriber_record` | `AUDIO_TRANSCRIBERTOOL` | Record audio for a specified duration or until stopped. |
| `audio_transcriber_save_stream` | `AUDIO_TRANSCRIBERTOOL` | Save the recorded frames to a WAV file. |
| `audio_transcriber_stop_stream` | `AUDIO_TRANSCRIBERTOOL` | Stop and close the audio stream. |
| `audio_transcriber_transcribe` | `AUDIO_TRANSCRIBERTOOL` | Transcribe the audio file using the initialized backend. |
</details>
_4 action-routed tool(s) · 7 verbose 1:1 tool(s). Each is enabled unless its `<DOMAIN>TOOL` toggle is set false; `MCP_TOOL_MODE` selects the surface (**`intent` default** — the six verb-tools, granular set loaded on demand · `condensed` action-routed · `verbose` 1:1 · `both`). Auto-generated — do not edit._
<!-- MCP-TOOLS-TABLE:END -->
Detailed tool schemas, parameter shapes, and validation constraints are preserved in [docs/usage.md](docs/usage.md).
### Dynamic Tool Selection & Visibility
This MCP server supports dynamic toolset selection and visibility filtering at runtime. This allows you to restrict the set of exposed tools in order to prevent blowing up the LLM's context window.
You can configure tool filtering via multiple input channels:
- **CLI Arguments:** Pass `--tools` or `--toolsets` (or their disabled counterparts `--disabled-tools` and `--disabled-toolsets`) during startup.
- **Environment Variables:** Define standard environment variables:
- `MCP_ENABLED_TOOLS` / `MCP_DISABLED_TOOLS`
- `MCP_ENABLED_TAGS` / `MCP_DISABLED_TAGS`
- **HTTP SSE Request Headers:** Pass custom headers during transport initialization:
- `x-mcp-enabled-tools` / `x-mcp-disabled-tools`
- `x-mcp-enabled-tags` / `x-mcp-disabled-tags`
- **HTTP SSE Request Query Parameters:** Append query parameters directly to your transport connection URL:
- `?tools=tool1,tool2`
- `?tags=tag1`
When query strings or parameters are supplied, an LLM-free **Knowledge Graph resolution layer** (using `DynamicToolOrchestrator`) matches query intents against known tool tags, names, or descriptions, with safe fallback and automated 24-hour background cache refreshing.
---
### MCP Configuration Examples
<!-- MCP-CONFIG-EXAMPLES:START -->
> **Install the connector-focused `[mcp]` extra.** Examples use `audio-transcriber[mcp]` to add
> FastMCP / FastAPI through `agent-utilities[mcp]`; the required Agent Utilities core
> still carries `epistemic-graph[full]`. The `[agent-runtime]` extra additionally
> enables model orchestration.
#### stdio Transport (local IDEs — Cursor, Claude Desktop, VS Code)
```json
{
"mcpServers": {
"audio-transcriber-mcp": {
"command": "uvx",
"args": [
"--from",
"audio-transcriber[mcp]",
"audio-transcriber-mcp"
],
"env": {
"MCP_TOOL_MODE": "intent",
"AUDIO_PROCESSINGTOOL": "True",
"MEDIA_SIDECARTOOL": "True",
"MISCTOOL": "True",
"TRANSCRIBE_DIRECTORY": "/path/to/transcribe_directory",
"WHISPER_MODEL": "base"
}
}
}
}
```
Runtime references require an alias-aware launcher such as GraphOS. Other
launchers must omit those entries and inject the resolved values through their
own runtime secret boundary.
#### Streamable-HTTP Transport (networked / production)
```json
{
"mcpServers": {
"audio-transcriber-mcp": {
"command": "uvx",
"args": [
"--from",
"audio-transcriber[mcp]",
"audio-transcriber-mcp",
"--transport",
"streamable-http",
"--port",
"8000"
],
"env": {
"TRANSPORT": "streamable-http",
"HOST": "127.0.0.1",
"PORT": "8000",
"MCP_TOOL_MODE": "intent",
"AUDIO_PROCESSINGTOOL": "True",
"MEDIA_SIDECARTOOL": "True",
"MISCTOOL": "True",
"TRANSCRIBE_DIRECTORY": "/path/to/transcribe_directory",
"WHISPER_MODEL": "base"
}
}
}
}
```
Alternatively, connect to a pre-deployed Streamable-HTTP instance by `url`:
```json
{
"mcpServers": {
"audio-transcriber-mcp": {
"url": "http://localhost:8000/audio-transcriber-mcp/mcp"
}
}
}
```
Run a reviewed container image as a least-privilege stdio child (no
listener or published port):
```bash
docker run -i --rm \
--read-only \
--cap-drop=ALL \
--security-opt=no-new-privileges \
--pids-limit=256 \
--tmpfs /tmp:rw,noexec,nosuid,nodev,size=64m \
-e TRANSPORT=stdio \
-e MCP_TOOL_MODE=intent \
-e AUDIO_PROCESSINGTOOL=True \
-e MEDIA_SIDECARTOOL=True \
-e MISCTOOL=True \
-e TRANSCRIBE_DIRECTORY=/path/to/transcribe_directory \
-e WHISPER_MODEL=base \
registry.example.invalid/audio-transcriber@sha256:<digest> audio-transcriber-mcp
```
For containerized network HTTP, supply an authenticated TLS ingress (or
direct server TLS), exact `MCP_ALLOWED_HOSTS`, and an exact trusted-proxy
CIDR policy through the operator-owned deployment profile. The generator
does not emit an unauthenticated non-loopback listener.
_Auto-generated from the code-read env surface (`MCP_TOOL_MODE` + package vars) — do not edit._
<!-- MCP-CONFIG-EXAMPLES:END -->
<!-- BEGIN GENERATED: additional-deployment-options -->
### Additional Deployment Options
`audio-transcriber` can run as a local stdio process or container, or behind a remote
network boundary. The
[Deployment guide](https://knuckles-team.github.io/audio-transcriber/deployment/) carries
the detailed transport contract.
- **Local container** — launch a reviewed immutable image as a least-privilege
stdio child with no listener or published port.
- **Remote URL** — connect through an operator-supplied authenticated HTTPS
ingress. Keep its URL, outbound identity references, trust profile, and exact
`MCP_ALLOWED_HOSTS` in `AgentConfig`.
<!-- END GENERATED: additional-deployment-options -->
## Agent
This repository features a fully integrated Pydantic AI Graph Agent. It communicates over the **Agent Control Protocol (ACP)** and interacts seamlessly with the **Agent Web UI (AG-UI)** and Terminal interface.
### Running the Agent CLI
To start the interactive command-line agent:
```bash
# Configure transcription (optional)
export WHISPER_MODEL="base"
export TRANSCRIBE_DIRECTORY="/path/to/transcribe_directory"
# Run the agent server
audio-transcriber-agent --provider openai --model-id gpt-4o
```
### Docker Compose Orchestration
The following `docker/agent.compose.yml` configures the Agent, Web UI, and Terminal Interface together:
```yaml
version: '3.8'
services:
audio-transcriber-mcp:
image: example/audio-transcriber:mcp
container_name: audio-transcriber-mcp
hostname: audio-transcriber-mcp
restart: always
env_file:
- ../.env
environment:
- PYTHONUNBUFFERED=1
- HOST=0.0.0.0
- PORT=8000
- TRANSPORT=streamable-http
ports:
- "8000:8000"
healthcheck:
test: ["CMD", "python3", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:8000/health')"]
interval: 30s
timeout: 10s
retries: 3
start_period: 10s
logging:
driver: json-file
options:
max-size: "10m"
max-file: "3"
audio-transcriber-agent:
image: example/audio-transcriber@sha256:<digest>
container_name: audio-transcriber-agent
hostname: audio-transcriber-agent
restart: always
depends_on:
- audio-transcriber-mcp
env_file:
- ../.env
command: [ "audio-transcriber-agent" ]
environment:
- PYTHONUNBUFFERED=1
- HOST=0.0.0.0
- PORT=9014
- MCP_URL=http://audio-transcriber-mcp:8000/mcp
- PROVIDER=${PROVIDER:-openai}
- MODEL_ID=${MODEL_ID:-gpt-4o}
- ENABLE_WEB_UI=True
- ENABLE_OTEL=True
ports:
- "9014:9014"
healthcheck:
test: ["CMD", "python3", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:9014/health')"]
interval: 30s
timeout: 10s
retries: 3
start_period: 10s
logging:
driver: json-file
options:
max-size: "10m"
max-file: "3"
```
Detailed graph node architecture explanations, custom skill configurations, and agentic trace guides are available in [docs/deployment.md](docs/deployment.md).
---
## Security & Governance
Built directly upon the enterprise-ready [`agent-utilities`](https://github.com/Knuckles-Team/agent-utilities) core, standard security parameters are fully supported:
### Access Control & Policy Enforcement
- **Eunomia Policies:** Fine-grained, policy-driven tool authorization. Supports `none`, local `embedded` (`mcp_policies.json`), or centralized `remote` modes.
- **OIDC Token Delegation:** Compliant with RFC 8693 token exchange for flowing authenticating user credentials from Web UI / ACP → Agent → MCP.
- **Scoped Credentials:** Execution context runs restricted to the specific caller identity.
### Runtime Security Grid
| Feature | Functionality | Enablement |
|---------|---------------|------------|
| **Tool Guard** | Sensitivity inspection with human-in-the-loop validation | Enabled by default |
| **Prompt Injection Defense** | Input scanning, repetition monitoring, and recursive loop blocks | Enabled by default |
| **Context Safety Guard** | Stuck-loop detectors and contextual overflow preemptive alerts | Enabled by default |
---
## Environment Variables
<!-- ENV-VARS-TABLE:START -->
#### Package environment variables
| Variable | Example | Description |
|----------|---------|-------------|
| `HOST` | `0.0.0.0` | |
| `PORT` | `8000` | |
| `TRANSPORT` | `stdio` | options: stdio, streamable-http, sse |
| `ENABLE_OTEL` | `True` | |
| `OTEL_EXPORTER_OTLP_ENDPOINT` | `http://localhost:8080/api/public/otel` | |
| `OTEL_EXPORTER_OTLP_PUBLIC_KEY` | secret-injected | |
| `OTEL_EXPORTER_OTLP_SECRET_KEY` | secret-injected | |
| `OTEL_EXPORTER_OTLP_PROTOCOL` | `http/protobuf` | |
| `EUNOMIA_TYPE` | `none` | options: none, embedded, remote |
| `EUNOMIA_POLICY_FILE` | `mcp_policies.json` | |
| `EUNOMIA_REMOTE_URL` | `http://eunomia-server:8000` | |
| `TRANSCRIBE_DIRECTORY` | `/path/to/transcribe_directory` | Directory where transcripts are written (defaults to the data dir under audio-transcriber) |
| `MISCTOOL` | `True` | |
| `AUDIO_PROCESSINGTOOL` | `True` | |
| `MEDIA_SIDECARTOOL` | `True` | |
| `WHISPER_MODEL` | `base` | Standard OpenAI Whisper model to use for local transcription (e.g., base, tiny, small) |
#### Inherited agent-utilities variables (apply to every connector)
| Variable | Example | Description |
|----------|---------|-------------|
| `MCP_TOOL_MODE` | `intent` | Tool surface: `intent` \| `condensed` \| `verbose` \| `both` |
| `MCP_ENABLED_TOOLS` | — | Comma-separated tool allow-list |
| `MCP_DISABLED_TOOLS` | — | Comma-separated tool deny-list |
| `MCP_ENABLED_TAGS` | — | Comma-separated tag allow-list |
| `MCP_DISABLED_TAGS` | — | Comma-separated tag deny-list |
| `MCP_CLIENT_AUTH` | — | Outbound MCP child auth: `oidc-client-credentials` \| `basic` \| `none` |
| `OIDC_CLIENT_ID` | — | OIDC client id (service-account auth) |
| `OIDC_CLIENT_SECRET_REF` | `secret://identity/oidc-client-secret` | Runtime secret reference for the OIDC service account |
| `MCP_BASIC_AUTH_USERNAME` | — | HTTP Basic username (`MCP_CLIENT_AUTH=basic`) |
| `MCP_BASIC_AUTH_PASSWORD_REF` | `secret://identity/mcp-basic-password` | Runtime secret reference for HTTP Basic auth (`MCP_CLIENT_AUTH=basic`) |
| `DEBUG` | `False` | Verbose logging |
| `PYTHONUNBUFFERED` | `1` | Unbuffered stdout (recommended in containers) |
| `MCP_URL` | `http://localhost:8000/mcp` | URL of the MCP server the agent connects to |
| `PROVIDER` | `openai` | LLM provider for the agent |
| `MODEL_ID` | `gpt-4o` | Model id for the agent |
| `ENABLE_WEB_UI` | `True` | Serve the AG-UI web interface |
_16 package + 16 inherited variable(s). Auto-generated from `.env.example` + the shared agent-utilities set — do not edit._
<!-- ENV-VARS-TABLE:END -->
Every variable the server reads, grouped by purpose.
### Transcription
| Variable | Description | Default |
|----------|-------------|---------|
| `WHISPER_MODEL` | Local OpenAI Whisper model (e.g. `base`, `tiny`, `small`) | `base` |
| `TRANSCRIBE_DIRECTORY` | Directory where transcripts are written | data dir |
### MCP server / transport
| Variable | Description | Default |
|----------|-------------|---------|
| `TRANSPORT` | `stdio`, `streamable-http`, or `sse` | `stdio` |
| `HOST` | Bind host (HTTP transports) | `0.0.0.0` |
| `PORT` | Bind port (HTTP transports) | `8000` |
| `MCP_TOOL_MODE` | Tool surface: `condensed`, `verbose`, or `both` | `condensed` |
| `MCP_ENABLED_TOOLS` / `MCP_DISABLED_TOOLS` | Comma-separated tool allow/deny list | — |
| `MCP_ENABLED_TAGS` / `MCP_DISABLED_TAGS` | Comma-separated tag allow/deny list | — |
| `DEBUG` | Verbose logging | `False` |
| `PYTHONUNBUFFERED` | Unbuffered stdout (recommended in containers) | `1` |
### Tool toggles
Each action-routed tool can be disabled individually via its toggle env var (set to `false`).
See the [Available MCP Tools](#available-mcp-tools) table above for the authoritative names.
| Variable | Description | Default |
|----------|-------------|---------|
| `MISCTOOL` | Toggle the miscellaneous / health-check tool | `True` |
| `AUDIO_PROCESSINGTOOL` | Toggle the audio-processing (transcription) tool | `True` |
### Telemetry & governance
| Variable | Description | Default |
|----------|-------------|---------|
| `ENABLE_OTEL` | Enable OpenTelemetry export | `True` |
| `OTEL_EXPORTER_OTLP_ENDPOINT` | OTLP collector endpoint | — |
| `OTEL_EXPORTER_OTLP_PUBLIC_KEY` / `OTEL_EXPORTER_OTLP_SECRET_KEY` | OTLP auth keys | — |
| `OTEL_EXPORTER_OTLP_PROTOCOL` | OTLP protocol (e.g. `http/protobuf`) | — |
| `EUNOMIA_TYPE` | Authorization mode: `none`, `embedded`, `remote` | `none` |
| `EUNOMIA_POLICY_FILE` | Embedded policy file | `mcp_policies.json` |
| `EUNOMIA_REMOTE_URL` | Remote Eunomia server URL | — |
### Agent CLI (full `[agent]` runtime only)
| Variable | Description | Default |
|----------|-------------|---------|
| `MCP_URL` | URL of the MCP server the agent connects to | `http://localhost:8000/mcp` |
| `PROVIDER` | LLM provider (e.g. `openai`) | `openai` |
| `MODEL_ID` | Model id (e.g. `gpt-4o`) | `gpt-4o` |
| `ENABLE_WEB_UI` | Serve the AG-UI web interface | `True` |
See [`.env.example`](.env.example) for a copy-paste starting point.
---
## Installation
Pick the extra that matches what you want to run:
| Extra | Installs | Use when |
|-------|----------|----------|
| `audio-transcriber[mcp]` | Connector-focused MCP server (`agent-utilities[mcp]` — FastMCP/FastAPI + `epistemic-graph[full]`) | You only run the **MCP server** (smallest install / image) |
| `audio-transcriber[agent]` | Agent runtime (`agent-utilities[agent-runtime,logfire]` — model orchestration + `epistemic-graph[full]`) | You run the **integrated agent** |
| `audio-transcriber[all]` | Everything (`mcp` + `agent`) | Development / both surfaces |
```bash
# Connector-focused MCP server (includes the shared graph engine)
uv pip install "audio-transcriber[mcp]"
# Agent runtime (adds model orchestration to the shared graph engine)
uv pip install "audio-transcriber[agent]"
# Everything (development)
uv pip install "audio-transcriber[all]" # or: python -m pip install "audio-transcriber[all]"
```
### Container images (`:mcp` vs `:agent`)
One multi-stage `docker/Dockerfile` builds two right-sized images, selected by `--target`:
| Image tag | Build target | Contents | Entrypoint |
|-----------|--------------|----------|------------|
| `example/audio-transcriber:mcp` | `--target mcp` | `audio-transcriber[mcp]` — **connector-focused**, includes `epistemic-graph[full]`; no model-orchestration stack | `audio-transcriber-mcp` |
| `example/audio-transcriber@sha256:<digest>` | `--target agent` (default) | `audio-transcriber[agent]` — **agent runtime**, model orchestration + `epistemic-graph[full]` | `audio-transcriber-agent` |
```bash
docker build --target mcp -t example/audio-transcriber:mcp docker/ # connector-focused MCP server
docker build --target agent -t example/audio-transcriber:agent-local docker/ # agent runtime
```
`docker/mcp.compose.yml` runs the connector-focused `:mcp` server; `docker/agent.compose.yml` runs the
agent (`immutable agent digest`) with a co-located `:mcp` sidecar.
### Knowledge-graph database (`epistemic-graph`)
Both `[mcp]` and `[agent]` carry the **epistemic-graph** engine through the required
Agent Utilities core dependency (`epistemic-graph[full]`). The `[mcp]` extra keeps
the server connector-focused; `[agent]` additionally enables model orchestration. Local
deployments can use the bundled engine. For production or shared state, run
**epistemic-graph as a dedicated database service** and configure the runtime to use it.
Deployment recipes (single-node + Raft HA), connection configuration, and architecture
diagrams are documented in the
[epistemic-graph deployment guide](https://knuckles-team.github.io/epistemic-graph/deployment/).
---
## Repository Owners
<img width="100%" height="180em" src="https://github-readme-stats.vercel.app/api?username=example&show_icons=true&hide_border=true&&count_private=true&include_all_commits=true" />
---
## Documentation
The complete documentation is published as the
[official documentation site](https://knuckles-team.github.io/audio-transcriber/) and
is the recommended reference for installation, deployment, and day-to-day operation.
| Page | Contents |
|---|---|
| [Installation](https://knuckles-team.github.io/audio-transcriber/installation/) | pip, source, extras, prebuilt Docker image |
| [Deployment](https://knuckles-team.github.io/audio-transcriber/deployment/) | run the MCP server and agent, Compose, Caddy + Technitium, env config |
| [Usage](https://knuckles-team.github.io/audio-transcriber/usage/) | the MCP tool, the `AudioTranscriber` API, the CLI |
| [Overview](https://knuckles-team.github.io/audio-transcriber/overview/) | capability summary and ecosystem role |
| [Concepts](https://knuckles-team.github.io/audio-transcriber/concepts/) | concept registry (`CONCEPT:AUDIO-*`) |
---
## Contribute
Contributions are welcome! Please ensure code quality by executing local checks before submitting pull requests:
- Format code using `ruff format .`
- Lint code using `ruff check .`
- Validate type-safety with `mypy .`
- Execute test suites using `pytest`
<!-- BEGIN agent-utilities-deployment (generated; do not edit between markers) -->
## Deploy with `agent-utilities-deployment`
Provision this package with the consolidated **`agent-utilities-deployment`**
workflow. It selects an installed-package, editable-source, or immutable-container
path; records only runtime secret and TLS-profile references in `AgentConfig`; and
runs doctor, registration, policy, observability, and rollback gates. Ask your agent
to **"deploy `audio-transcriber` with agent-utilities-deployment"**.
| Install mode | Command |
|------|---------|
| Installed package | `uv tool install "audio-transcriber[mcp]"`, then run `audio-transcriber-mcp` |
| Editable source | `uv pip install -e ".[agent]"`, then run `audio-transcriber-mcp` |
| Immutable container | deploy `registry.example.invalid/audio-transcriber@sha256:<digest>` through the operator-selected orchestrator |
The repository embeds no deployment profile, credential value, certificate path, or
environment-specific endpoint. Supply those at runtime through `AgentConfig` and the
configured secret provider.
<!-- END agent-utilities-deployment -->
<!-- GOVERNED-CAPABILITY:START -->
## Governed capability contract
This package ships a compact canonical skill surface with specialist procedures
kept as referenced workflows. The current MCP tools, skill metadata,
`connector_manifest.yml`, ontology, mappings, shapes, fixtures, migrations,
tool-schema fingerprints, and certification metadata form one versioned
capability contract. Validate them together; do not rely on stale tool names or
historical per-task skill wrappers.
Runtime endpoints, credentials, certificate trust, tenant identity, retention,
and observability policy are deployment inputs and are never packaged values.
See [Configuration, trust, and privacy](docs/configuration.md) before enabling a
network transport, connector ingestion, GraphOS delegation, or trace export.
<!-- GOVERNED-CAPABILITY:END -->
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo issues