CtrlK
BlogDocsLog inGet started
Tessl Logo

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".

95

1.19x
Quality

Does it follow best practices?

Impact

100%

1.19x

Average score across 3 eval scenarios

SecuritybySnyk

Advisory

Suggest reviewing before use

SKILL.md
Quality
Evals
Security

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An actionable, token-efficient skill body with executable sync/async examples and dense SDK-specific gotchas. It loses points on workflow clarity (no validation feedback loop) and progressive disclosure (bundled reference files are referenced but not present in the bundle).

Suggestions

Add an explicit validate/recover step for the patched socket_client.py workflow (e.g., a check after regeneration that the manual patch and .fernignore still hold) to lift workflow_clarity to 3.

Either ship the referenced reference.md in a references/ bundle (one level deep, clearly signaled) or remove the in-repo reference pointer in favor of the external links so progressive disclosure matches the actual bundle structure.

Clarify which event types signal a recoverable vs terminal error (ListenV2FatalError vs ListenV2ConfigureFailure) so Claude can implement a retry/feedback loop on connection errors.

DimensionReasoningScore

Conciseness

Lean throughout: no preamble explaining what STT or websockets are; every section (Quick start, Key parameters, Events, Gotchas) earns its tokens with SDK-specific, non-obvious detail like sample_rate being a string and the patched socket_client.py.

3 / 3

Actionability

Fully executable Quick start (sync) and Async equivalent code blocks with real imports, plus a concrete parameters table and a numbered Gotchas list of copy-paste-ready rules ("Close with send_close_stream(...)", "Send ~80ms audio chunks").

3 / 3

Workflow Clarity

A clear connect -> handle events -> stream -> close sequence is shown, but there is no explicit validation/verification checkpoint or error-recovery feedback loop; Gotcha #8 flags a fragile manual patch but offers no verify step.

2 / 3

Progressive Disclosure

Well-organized into sections, but no bundle files exist (references/, scripts/, assets/ absent) despite the body pointing to reference.md, AsyncAPI/Context7 links, and example files; references are signaled but the in-repo reference.md is not present alongside.

2 / 3

Total

10

/

12

Passed

Description

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.

A highly specific, well-scoped description that answers what and when with concrete API references, natural trigger terms, and explicit disambiguation against adjacent skills. It is third person and free of fluff.

DimensionReasoningScore

Specificity

Lists multiple concrete, specific actions ("client.listen.v2.connect", Flux models, end-of-turn detection, EOT) rather than vague language.

3 / 3

Completeness

Explicitly answers both what (turn-aware streaming transcription at /v2/listen, Flux models, EOT detection) and when ("Use when writing or reviewing Python code... that calls Deepgram Conversational STT v2 / Flux") with explicit trigger guidance.

3 / 3

Trigger Term Quality

Excellent coverage of natural trigger phrases a user would say ("flux", "v2 listen", "conversational STT", "turn detection", "end of turn", "EOT", "listen.v2", model IDs).

3 / 3

Distinctiveness Conflict Risk

Clearly distinguished from sibling skills via explicit routing rules (use deepgram-python-speech-to-text for v1 ASR, deepgram-python-voice-agent for full-duplex assistants), making misfires unlikely.

3 / 3

Total

12

/

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
deepgram/deepgram-python-sdk
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.