audiocraft-audio-generation

PyTorch library for audio generation including text-to-music (MusicGen) and text-to-sound (AudioGen). Use when you need to generate music from text descriptions, create sound effects, or perform melody-conditioned music generation.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/coding/audiocraft/SKILL.md

Quality

Content

53%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 content is highly actionable with copy-paste code across the main AudioCraft workflows and correctly uses two reference files for advanced/troubleshooting material. Its weaknesses are verbosity — much of the overview duplicates library knowledge and belongs in references — and missing validation/verification checkpoints in the batch-generation workflows.

Suggestions

Trim the in-file overview: move the model-variant, generation-parameter, and VRAM tables plus the MusicGen-Style and EnCodec sections into references/advanced-usage.md, leaving a lean quick-start plus links.

Add explicit validation/verification steps to the batch-generation and Gradio workflows (e.g. assert wav is non-empty, check output file exists and sample rate, catch CUDA OOM and retry with a smaller model) to lift workflow clarity above the batch-operation cap of 3.

De-duplicate the repeated load/configure/generate/save code pattern by consolidating near-identical MusicGen snippets into one base example plus a table of variant-specific differences.

Dimension	Reasoning	Score
Conciseness	The 555-line body extensively restates library knowledge Claude already has (model variant tables, parameter tables, architecture diagrams, repeated near-identical generation snippets across sections) and pads the in-file overview with material that belongs in the references, matching the "noticeably verbose; several padded sections" anchor.	2 / 5
Actionability	Mostly executable copy-paste-ready code covers the common cases (text-to-music, melody, stereo, AudioGen, EnCodec) with real imports and save calls, but several snippets omit validation/error handling and some (e.g. continuation, style-only) are incomplete, leaving minor gaps.	4 / 5
Workflow Clarity	Sections are sequenced by capability and the batch-processing and Gradio workflows show steps, but batch/destructive-style generation workflows lack explicit validation checkpoints or verify-after-generate feedback loops, which the rubric caps at 3 for such operations.	3 / 5
Progressive Disclosure	Two real bundle files exist (references/advanced-usage.md, references/troubleshooting.md) and are clearly linked from a References section, with quick-start inline and detail offloaded; however the body still inlines a large amount of reference-grade content (full model/parameter/VRAM tables, style and EnCodec sections) that could live in the advanced file.	4 / 5
	Total	13 / 20 Passed

Description

92%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, specific description that names concrete capabilities and provides an explicit "Use when" trigger covering the main tasks. The only minor gap is the absence of synonym/extension variants that would push trigger-term coverage to fully comprehensive.

Dimension	Reasoning	Score
Specificity	Names the domain and lists multiple concrete capabilities — "text-to-music (MusicGen)", "text-to-sound (AudioGen)", "generate music from text descriptions", "create sound effects", "melody-conditioned music generation" — giving comprehensive coverage of the library's actions.	5 / 5
Completeness	It clearly answers "what" (PyTorch library for audio generation including MusicGen and AudioGen) and "when" via an explicit "Use when you need to generate music from text descriptions, create sound effects, or perform melody-conditioned music generation" clause with concrete trigger phrases.	5 / 5
Trigger Term Quality	Good natural keywords — "generate music from text descriptions", "create sound effects", "text-to-music" — but the anchor 5 example expects synonyms and file extensions (e.g. .wav, audio files), which are absent, so it falls just below comprehensive.	4 / 5
Distinctiveness Conflict Risk	It occupies a clear niche (AudioCraft text-to-music/sound via MusicGen/AudioGen) with distinct triggers unlikely to fire for unrelated skills; minimal conflict risk.	5 / 5
	Total	19 / 20 Passed

Validation

87%

Warnings & errors only

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

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (566 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/coding/audiocraft/SKILL.md
Commit: fcd2e11

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.