CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

deepgram-java-conversational-stt

Use when writing or reviewing Java code in this repo that calls Deepgram Conversational STT v2 / Flux over `/v2/listen`. Covers `client.listen().v2().v2WebSocket()`, `V2ConnectOptions`, `onTurnInfo`, and turn-aware close handling. Use `deepgram-java-speech-to-text` for standard v1 transcription and `deepgram-java-voice-agent` for fully interactive assistants. Triggers include "flux", "conversational stt", "listen v2", "turn detection", "end of turn", and "eot".

72

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

78%Weight 40%Scale 1-5

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

A strong, dense body: executable Java quick start, verified API surface, and genuinely useful gotchas (plain-String model, v2-specific close, no language_hint) with zero padding. The main gaps are that the audio-streaming core is only a commented-out sendMedia line, there is no error-handling feedback loop, and a small promotional tail section slightly dilutes token efficiency.

Suggestions

Replace the commented-out `// wsClient.sendMedia(okio.ByteString.of(audioChunk));` with a short real loop (e.g., reading chunks from an InputStream/Source and calling sendMedia), since streaming audio is the skill's core task and is currently never shown end-to-end.

Wire `onErrorMessage(...)` (and the generic error hooks) into the Quick start or a short 'Error handling' note so the workflow includes an explicit recovery checkpoint instead of only listing the handler in the API surface.

Merge 'Example files in this repo' into the 'API reference (layered)' section (which already cites `examples/listen/LiveStreamingV2.java`) and trim or shorten the 'Central product skills' promotional section to improve token efficiency.

DimensionReasoningScore

Conciseness

The body is lean and assumes competence — no explanations of what STT or WebSockets are, dense gotchas and parameter lists — but the closing 'Central product skills' promo section and a separate 'Example files in this repo' section (repeating `examples/listen/LiveStreamingV2.java` already cited under 'API reference (layered)') are minor tokens that could be trimmed, fitting the 'efficient; minor instances of over-explanation' anchor rather than 'every token earns its place'.

4 / 5

Actionability

The Quick start is real, compilable Java with imports, handler wiring, connect-with-timeout, and close flow, and 'Key parameters' lists verified connect options — but the code depends on `client` from the earlier Authentication section and the core streaming action `sendMedia(...)` appears only as a commented-out line, so the actual audio-feeding loop is never demonstrated (acknowledged by Gotcha 6). Mostly executable with minor gaps, not fully copy-paste ready.

4 / 5

Workflow Clarity

A clear implicit sequence (auth → client → handlers → connect(...).get(...) → media → sendCloseStream) with an explicit checkpoint in Gotcha 7 ('Wait for `connect(...).get(...)` before sending media'), but no error-recovery feedback loop — `onErrorMessage` is listed in the API surface yet never wired into the flow — which keeps it below the explicit-validation anchor. The skill is non-destructive, so the batch/destruction cap at 3 does not apply.

4 / 5

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ are absent) and none are needed; the body is a compact, well-sectioned overview with well-signaled, one-level-deep, priority-ordered references ('API reference (layered)': in-repo source → AsyncAPI → Context7 → product docs, honestly noting 'No `reference.md` exists in this checkout'), and nothing bulky is inlined that belongs in a separate file.

5 / 5

Total

17

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is exemplary: concrete capabilities, an explicit 'Use when' clause with natural trigger terms including synonyms, and explicit disambiguation from two sibling Deepgram skills. It reads like the reference 'good' examples — nothing vague, padded, or over-claimed.

DimensionReasoningScore

Specificity

Names concrete, specific capabilities — 'writing or reviewing Java code in this repo that calls Deepgram Conversational STT v2 / Flux over `/v2/listen`' plus a precise API surface ('client.listen().v2().v2WebSocket()', 'V2ConnectOptions', 'onTurnInfo', 'turn-aware close handling') — with no vague filler and neutral voice matching the reference good examples.

5 / 5

Completeness

Explicitly answers both 'what' (covers v2WebSocket(), V2ConnectOptions, onTurnInfo, turn-aware close handling over /v2/listen) and 'when' ('Use when writing or reviewing Java code in this repo...') with concrete trigger phrases, matching the top anchor exactly.

5 / 5

Trigger Term Quality

Lists six natural trigger terms a user would actually say ('flux', 'conversational stt', 'listen v2', 'turn detection', 'end of turn', 'eot'), including the synonym pair 'end of turn'/'eot'; no obviously missing natural variant within this skill's niche, so it fits the comprehensive anchor rather than the 'a few natural terms missing' anchor.

5 / 5

Distinctiveness Conflict Risk

Actively de-conflicts against sibling skills ('Use `deepgram-java-speech-to-text` for standard v1 transcription and `deepgram-java-voice-agent` for fully interactive assistants'), giving it a clear niche with distinct triggers and minimal conflict risk.

5 / 5

Total

20

/

20

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

deepgram/deepgram-java-sdk

Path

.agents/skills/deepgram-java-conversational-stt/SKILL.md

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.