deepgram-python-conversational-stt
Use when writing or reviewing Python code in this repo that calls Deepgram Conversational STT v2 / Flux (`/v2/listen`) for turn-aware streaming transcription. Covers `client.listen.v2.connect(...)`, Flux models, end-of-turn detection. Use `deepgram-python-speech-to-text` for standard v1 ASR, `deepgram-python-voice-agent` for full-duplex interactive assistants. Triggers include "flux", "v2 listen", "conversational STT", "turn detection", "end of turn", "EOT", "listen.v2", "flux-general-en", "flux-general-multi".
88
82%
Does it follow best practices?
Impact
100%
1.19xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that hits all the marks: it specifies concrete actions and API details, provides an explicit 'Use when' clause with rich trigger terms, and proactively disambiguates from related skills to minimize conflict risk. The description is concise yet comprehensive, making it easy for Claude to select the right skill from a large pool.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and artifacts: 'client.listen.v2.connect(...)', Flux models, end-of-turn detection, '/v2/listen' endpoint, turn-aware streaming transcription. Very concrete and technical. | 3 / 3 |
Completeness | Clearly answers both 'what' (writing/reviewing Python code calling Deepgram Conversational STT v2/Flux for turn-aware streaming transcription) and 'when' (explicit 'Use when' clause at the start, plus explicit trigger terms). Also helpfully disambiguates from related skills. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms explicitly listed: 'flux', 'v2 listen', 'conversational STT', 'turn detection', 'end of turn', 'EOT', 'listen.v2', 'flux-general-en', 'flux-general-multi'. These are terms users would naturally use when working with this specific API. | 3 / 3 |
Distinctiveness Conflict Risk | Exceptionally distinctive — explicitly differentiates from related skills ('deepgram-python-speech-to-text' for v1 ASR, 'deepgram-python-voice-agent' for full-duplex assistants), creating clear boundaries. The specific API version (v2), endpoint ('/v2/listen'), and model names make conflicts very unlikely. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, actionable skill with excellent concrete code examples and a very useful gotchas section that captures hard-won knowledge about SDK quirks. Its main weaknesses are moderate verbosity (promotional content, routing guidance that could be in frontmatter), and a lack of explicit error recovery/validation workflows for what is inherently a fragile WebSocket streaming operation.
Suggestions
Add an explicit error recovery workflow: what to do when ListenV2FatalError is received, how to handle reconnection, and how to verify the connection is healthy.
Trim the 'Central product skills' section and 'When to use this product' routing — the frontmatter description already covers routing, and the npx install block is promotional rather than instructional.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient but includes some unnecessary sections like the 'Central product skills' promo block and the 'When to use this product' section which could be trimmed. The 'Related skills' and cross-references add some bloat. However, the core technical content is well-condensed. | 2 / 3 |
Actionability | Provides fully executable code for both sync and async usage, concrete parameter tables with specific values, and a comprehensive gotchas section with exact method names and types. The quick start is copy-paste ready with real SDK imports and method calls. | 3 / 3 |
Workflow Clarity | The quick start demonstrates a clear connect → register handlers → send audio → close sequence, but there's no explicit validation or error recovery workflow. For a WebSocket-based streaming operation, there should be more guidance on handling reconnection, verifying connection state, or validating that turns are being received correctly. | 2 / 3 |
Progressive Disclosure | References to external docs (AsyncAPI, product docs, reference.md) and example files are well-signaled, but no bundle files are provided to verify reference.md exists. The skill itself is somewhat long with inline content that could be split (e.g., the full parameter table and events section could be in a reference file), though it's not egregiously monolithic. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- deepgram/deepgram-python-sdk
- Commit
e169f63
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.