Conversation
|
|
📝 WalkthroughWalkthroughThis PR adds a new Gnani speech-to-text plugin for LiveKit agents, including plugin infrastructure, STT implementation with API integration, comprehensive test coverage, documentation, and packaging configuration. Changes
Sequence Diagram(s)sequenceDiagram
participant Client as Client/Test
participant STT as Gnani STT Plugin
participant Audio as Audio Processor
participant HTTP as HTTP Client
participant API as Gnani API
Client->>STT: recognize(audio_buffer, language?)
STT->>Audio: collect frames & convert to WAV
Audio-->>STT: wav_bytes
STT->>STT: build multipart form<br/>(language_code, audio_file)
STT->>HTTP: POST with auth headers<br/>(api_key, org_id, user_id, request_id)
HTTP->>API: HTTP POST request
API-->>HTTP: JSON response<br/>(success, transcript)
HTTP-->>STT: response body
STT->>STT: validate response<br/>parse transcript
STT->>STT: create SpeechEvent<br/>(FINAL_TRANSCRIPT, SpeechData)
STT-->>Client: SpeechEvent
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
🧪 Generate unit tests (beta)
📜 Recent review detailsConfiguration used: Organization UI Review profile: CHILL Plan: Pro 📒 Files selected for processing (3)
🚧 Files skipped from review as they are similar to previous changes (1)
🧰 Additional context used📓 Path-based instructions (1)**/*.py📄 CodeRabbit inference engine (AGENTS.md)
Files:
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
🔇 Additional comments (8)
✏️ Tip: You can disable this entire section by setting Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Fix all issues with AI agents
In `@CONTRIBUTING.md`:
- Around line 95-96: Update the incorrect plugin module name
`livekit.plugins.inyaai` to `livekit.plugins.gnani` wherever referenced (e.g.,
the two lines in CONTRIBUTING.md and the import in
tests/test_plugin_inyaai_stt.py); change the import path in the test from `from
livekit.plugins.inyaai...` to `from livekit.plugins.gnani...` and update any
test identifiers or docstrings that reference "inyaai" so they match the gnani
package naming.
In `@tests/test_plugin_inyaai_stt.py`:
- Around line 146-161: The test test_recognize_api_error should expect the HTTP
error exception raised by STT.recognize, not the network error; change the
assertion to expect APIStatusError (the exception raised for non-200 responses
in livekit.plugins.gnani.stt.STT.recognize) instead of APIConnectionError so the
test matches the implementation that throws APIStatusError for status != 200.
🧹 Nitpick comments (6)
livekit-plugins/livekit-plugins-gnani/README.md (1)
48-71: Basic usage example has incomplete imports.The example references
VoiceAssistant,silero, andttswithout showing their imports. This could confuse users trying to copy the example.📝 Suggested fix for imports
```python from livekit.agents import AutoSubscribe, JobContext, WorkerOptions, cli, llm +from livekit.agents.pipeline import VoiceAssistant from livekit.plugins import gnani +from livekit.plugins import silero +from livekit.plugins import openai as ttstests/test_plugin_inyaai_stt.py (1)
146-175: MoveAPIConnectionErrorimport to module level.
APIConnectionErroris imported multiple times inside individual test methods (lines 148, 167, 181). Moving it to the top-level imports improves readability and follows Python best practices.♻️ Suggested fix
Add to the imports at the top of the file:
from livekit import rtc from livekit.agents import stt +from livekit.agents import APIConnectionError from livekit.agents.stt import SpeechData, SpeechEventType from livekit.plugins.inyaai import STTThen remove the local imports from each test method.
livekit-plugins/livekit-plugins-gnani/pyproject.toml (1)
25-37: Consider adding Python 3.13 classifier.The classifiers list Python 3.9 through 3.12. If Python 3.13 is supported (which is likely given
>=3.9.0requirement), consider adding the classifier for discoverability on PyPI.📝 Suggested addition
"Programming Language :: Python :: 3.11", "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", "Programming Language :: Python :: 3 :: Only",livekit-plugins/livekit-plugins-gnani/setup.py (1)
31-32: Use context manager for file reading.The file handle for
README.mdis not explicitly closed. While Python's garbage collector will eventually close it, using a context manager is the recommended practice.♻️ Suggested fix
+with open("README.md", encoding="utf-8") as readme_file: + long_description = readme_file.read() + setup( name="livekit-plugins-gnani", version=about["__version__"], description="Agent Framework plugin for Gnani (Vachana) Speech-to-Text API.", - long_description=open("README.md").read(), + long_description=long_description, long_description_content_type="text/markdown",livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/stt.py (2)
109-112: Guard against closed aiohttp sessions.
If a providedClientSessionis already closed, the current logic will reuse it and fail later with a less actionable error. Consider checking.closedand recreating (or raising explicitly).♻️ Proposed fix
- def _ensure_session(self) -> aiohttp.ClientSession: - if not self._session: - self._session = utils.http_context.http_session() - return self._session + def _ensure_session(self) -> aiohttp.ClientSession: + if self._session is None or self._session.closed: + self._session = utils.http_context.http_session() + return self._session
122-222: Normalize language before adding to form data.
IfGnaniLanguagesis a plain Enum,FormDatamay serialize the enum name rather than the language code. Coercing to.valuewhen present avoids incorrect payloads.♻️ Proposed fix
- form_data.add_field("language_code", target_language) + language_code = ( + target_language.value if hasattr(target_language, "value") else target_language + ) + form_data.add_field("language_code", language_code)
📜 Review details
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
⛔ Files ignored due to path filters (1)
uv.lockis excluded by!**/*.lock
📒 Files selected for processing (14)
CONTRIBUTING.mdlivekit-plugins/livekit-plugins-gnani/README.mdlivekit-plugins/livekit-plugins-gnani/livekit/__init__.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/__init__.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/__init__.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/log.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/models.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/py.typedlivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/stt.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/version.pylivekit-plugins/livekit-plugins-gnani/pyproject.tomllivekit-plugins/livekit-plugins-gnani/setup.pytests/docker-compose.ymltests/test_plugin_inyaai_stt.py
🧰 Additional context used
📓 Path-based instructions (1)
**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
**/*.py: Format code with ruff
Run ruff linter and auto-fix issues
Run mypy type checker in strict mode
Maintain line length of 100 characters maximum
Ensure Python 3.9+ compatibility
Use Google-style docstrings
Files:
livekit-plugins/livekit-plugins-gnani/livekit/plugins/__init__.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/models.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/version.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/stt.pytests/test_plugin_inyaai_stt.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/log.pylivekit-plugins/livekit-plugins-gnani/livekit/__init__.pylivekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/__init__.pylivekit-plugins/livekit-plugins-gnani/setup.py
🧠 Learnings (1)
📚 Learning: 2026-01-16T07:44:56.353Z
Learnt from: CR
Repo: livekit/agents PR: 0
File: AGENTS.md:0-0
Timestamp: 2026-01-16T07:44:56.353Z
Learning: Follow the Plugin System pattern where plugins in livekit-plugins/ are separate packages registered via the Plugin base class
Applied to files:
CONTRIBUTING.mdlivekit-plugins/livekit-plugins-gnani/livekit/__init__.py
🧬 Code graph analysis (2)
tests/test_plugin_inyaai_stt.py (4)
livekit-agents/livekit/agents/stt/stt.py (3)
SpeechData(53-61)SpeechEventType(32-49)SpeechEvent(70-74)tests/conftest.py (1)
job_process(19-22)livekit-agents/livekit/agents/_exceptions.py (1)
APIConnectionError(84-88)livekit-agents/livekit/agents/utils/http_context.py (1)
http_session(40-51)
livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/__init__.py (2)
livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/stt.py (1)
STT(46-242)livekit-agents/livekit/agents/plugin.py (2)
Plugin(13-56)register_plugin(31-36)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (3)
- GitHub Check: type-check (3.9)
- GitHub Check: type-check (3.13)
- GitHub Check: unit-tests
🔇 Additional comments (24)
livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/py.typed (1)
1-1: LGTM!Correct PEP 561 marker file for indicating type hint support in the package.
tests/docker-compose.yml (2)
63-65: LGTM!Environment variables for InyaAI credentials follow the established pattern used by other plugins in this compose file.
86-86: LGTM!The extra_hosts mapping for
api.vachana.aicorrectly routes the Gnani API through toxiproxy for integration testing, consistent with other service mappings.livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/log.py (1)
1-17: LGTM!Standard logging setup following the established pattern in other LiveKit plugins. The logger name correctly matches the module path.
livekit-plugins/livekit-plugins-gnani/livekit/__init__.py (1)
1-2: LGTM!Correct namespace package setup using
pkgutil.extend_path, allowing thelivekitnamespace to be shared across multiple plugin packages. This follows the established plugin system pattern. Based on learnings, plugins in livekit-plugins/ are separate packages, and this namespace approach supports that architecture.livekit-plugins/livekit-plugins-gnani/livekit/plugins/__init__.py (1)
1-2: LGTM!Correct namespace package initialization for
livekit.plugins, consistent with the parent namespace setup.livekit-plugins/livekit-plugins-gnani/README.md (1)
1-135: LGTM!The documentation is comprehensive, covering installation, features, supported languages, configuration options, and usage examples. The API details and limitations sections are particularly helpful for users evaluating this plugin.
tests/test_plugin_inyaai_stt.py (3)
26-41: LGTM!The
audio_bufferfixture correctly creates a 1-second audio frame with proper 16-bit PCM parameters (16kHz, mono). This provides a realistic test fixture for STT recognition tests.
106-125: LGTM!The recognition success test properly mocks the API response and verifies all expected fields of the
SpeechEventincluding type, alternatives count, text, language, and confidence.
227-245: LGTM!Good coverage of all supported languages including the code-switching variant (
en-IN,hi-IN). This ensures the STT accepts the full range of documented language options.livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/version.py (1)
1-15: LGTM!Simple and clean version module following standard Python packaging conventions.
livekit-plugins/livekit-plugins-gnani/pyproject.toml (1)
38-41: LGTM!Dependencies are appropriately specified with minimum versions, allowing flexibility for users while ensuring compatibility with the required features.
livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/models.py (1)
17-29: LGTM!The
GnaniLanguagestype alias provides good type safety and IDE auto-completion for supported language codes. The inline comments documenting each language code are helpful.livekit-plugins/livekit-plugins-gnani/setup.py (2)
22-25: Version loading pattern is acceptable.The
exec()pattern for loading version from a separate file is a common setuptools idiom that avoids import-time side effects. This is fine for packaging purposes.
37-42: LGTM!The namespace package discovery and dependencies are correctly configured, matching the
pyproject.tomlspecification for consistency.livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/__init__.py (4)
15-26: Clear module contract and exports.
Docstring and__all__make the public surface explicit and tidy.
34-36: Plugin initialization wiring looks good.
Constructor forwards name/version/package/logger cleanly.
39-39: Confirm import-time registration happens on the main thread.
Plugin.register_pluginraises outside the main thread; please verify this module is only imported on the main thread or registration is deferred to a controlled startup path.
41-48: Doc cleanup for unexported symbols is fine.
__pdoc__filtering aligns with the explicit__all__.livekit-plugins/livekit-plugins-gnani/livekit/plugins/gnani/stt.py (5)
41-43: Options dataclass is clean and minimal.
46-100: Credential validation and initialization look solid.
Clear env fallbacks and explicit errors.
101-107: Explicit model/provider identifiers are good.
114-120: Option updates are straightforward.
223-242: SpeechEvent assembly is clear.
✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.
3 participants
We have integrated Gnani's Speech-to-Text with LiveKit Agents. The service provides low-latency, high-accuracy transcription for Indian languages and accents, supporting English (Indian), Hindi, Tamil, Telugu, and more. It offers secure API key–based authentication and real-time transcription optimized for conversational and voice-based applications.
Summary by CodeRabbit
Release Notes
New Features
Documentation
Tests
✏️ Tip: You can customize this high-level summary in your review settings.