CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ax-audio

This skill helps an LLM generate correct audio code with @ax-llm/ax. Use when the user asks about ai.transcribe(), ai.speak(), signature audio inputs or outputs, agent audio behavior, .chat() conversational audio, OpenAI audio or realtime models, Gemini Live native audio, Grok Voice Agent models, voices, formats, transcripts, or how audio fits with structured outputs.

Quality

84%

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.

The body is a dense, highly actionable codegen rulebook with excellent executable examples and a clear upfront decision framework. Its weaknesses are inline time-sensitive model details and a monolithic single-file structure where provider-specific defaults and type references should be split out.

Suggestions

Split provider-specific sections (OpenAI Realtime defaults, Gemini Live, Grok Voice, Meta Muse) into reference files under references/ and keep a short per-provider summary with clearly signaled links in SKILL.md.

Move the full AxAudioFormat and AxSpeechConfig type definitions to a reference file, keeping only the most common values inline.

Isolate time-sensitive model currency notes (e.g., 'gpt-realtime-2.1 ... are newer reasoning models') into a dedicated models/versions section so they can be updated without touching the core rules.

DimensionReasoningScore

Conciseness

Nearly every line is a library-specific rule or executable code with no padding or explanation of known concepts, but time-sensitive model currency (e.g., 'gpt-realtime-2.1 ... are newer reasoning models', dated model names like 'gemini-3.8-flash-tts' and 'voxtral-mini-tts-2603') is woven inline rather than isolated in a dedicated section, which the guidelines penalize.

4 / 5

Actionability

Fully executable, copy-paste-ready TypeScript with imports covering batch transcribe/speak, signature artifacts, agent audio, .chat() audio, realtime WebSocket calls, and streaming, spanning the common cases for every supported provider.

5 / 5

Workflow Clarity

The opening decision guide ('Pick the smallest audio surface that matches the job') gives an unambiguous four-way selection mapped to later sections, and no destructive or batch operations require validation checkpoints; minor ambiguity remains between direct batch APIs and signature outputs.

4 / 5

Progressive Disclosure

No bundle files exist and all ~460 lines are inline, including full type definitions (AxAudioFormat, AxSpeechConfig), per-provider default lists, and lengthy Meta Muse streaming details that clearly belong in separate reference files; section headers provide structure, but content that should be separate is inline.

3 / 5

Total

16

/

20

Passed

Description

83%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.

A strong description that clearly states what the skill does and gives an explicit, trigger-rich 'Use when' clause in third person. Its main weaknesses are a few missing natural synonyms (TTS, speech-to-text) and provider-name triggers that slightly raise conflict risk with provider-specific skills.

DimensionReasoningScore

Specificity

The description names the domain ('generate correct audio code with @ax-llm/ax') and enumerates several concrete API surfaces ('ai.transcribe(), ai.speak(), signature audio inputs or outputs, agent audio behavior, .chat() conversational audio'), with minor gaps such as missing TTS/STT phrasing.

4 / 5

Completeness

It explicitly answers both 'what' ('helps an LLM generate correct audio code with @ax-llm/ax') and 'when' ('Use when the user asks about...') with a long list of concrete trigger phrases, matching the top anchor.

5 / 5

Trigger Term Quality

Good natural keyword coverage including 'voices, formats, transcripts', 'OpenAI audio or realtime models', and 'Gemini Live native audio', but common synonyms like 'TTS', 'text-to-speech', or 'speech-to-text' are absent, so it falls short of comprehensive.

4 / 5

Distinctiveness Conflict Risk

The audio-codegen niche for @ax-llm/ax is distinct, but broad provider trigger phrases like 'OpenAI audio or realtime models' create minor overlap risk with provider-specific skills.

4 / 5

Total

17

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

ax-llm/ax

Path

website/static/typescript/.well-known/agent-skills/ax-audio/SKILL.md

Revision

42fceae

Reviewed

Table of Contents

DiscoveryImplementationValidation

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.

Claim skill