whisper

Transcribe and translate speech in 99 languages.

Quality

59%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./optional-skills/mlops/whisper/SKILL.md

Quality

Content

65%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-ready code across the common transcription tasks, but it is padded with well-known background material, omits validation/error-recovery feedback loops for its batch operations, and fails to actually link its bundled references file. Tightening and proper progressive disclosure would raise quality significantly.

Suggestions

Replace the inline language list and model tables with a one-level reference link to references/languages.md so the bundled file is actually used and the overview stays lean.

Add explicit validation/error-recovery steps to the batch-processing and real-time sections (e.g. check transcription confidence, handle empty or corrupt audio, retry on failure).

Trim background content Claude already knows (GitHub star counts, explanatory comments, parameter counts) to improve token efficiency.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient with concrete code, but it pads with content Claude already knows (GitHub star counts, explanatory comments like '# Helps with:', '# Requires ffmpeg', full model-parameter tables) and could be tightened, matching the 'mostly efficient but some unnecessary explanation' anchor.	3 / 5
Actionability	Numerous sections provide fully executable, copy-paste-ready code and CLI commands covering common cases (basic transcription, language specification, translation, timestamps, batch processing), matching the 'fully executable, copy-paste ready' anchor.	5 / 5
Workflow Clarity	Recipes are well-sequenced, but the batch-processing loop and destructive/batch-style operations lack validation checkpoints and error-recovery feedback loops; per the feedback-loops note for batch operations this caps workflow_clarity at 3.	3 / 5
Progressive Disclosure	Sections are reasonably organized, but the available references/languages.md is never linked from the body and its language list is duplicated inline rather than deferred, matching the anchor for some structure with references present but not clearly signaled.	3 / 5
	Total	14 / 20 Passed

Description

53%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 clear and correctly voiced in third person, naming two concrete capabilities, but it omits any 'Use when...' trigger guidance and lacks the broader natural-language keywords users would actually say. It is competent but not comprehensive.

Suggestions

Add an explicit 'Use when...' clause naming concrete triggers, e.g. 'Use when transcribing podcasts, videos, or meetings, or translating audio to English.'

Broaden trigger terms to include natural phrases users say ('podcast', 'audio', 'video', 'subtitles', 'captions', 'speech-to-text').

Mention additional concrete actions (e.g. word-level timestamps, SRT/VTT subtitle export, batch transcription) to lift specificity toward comprehensive coverage.

Dimension	Reasoning	Score
Specificity	Names two concrete actions ('Transcribe and translate speech') but offers no further coverage, matching the anchor for naming the domain with 1-2 concrete actions without being comprehensive.	3 / 5
Completeness	The phrase 'Transcribe and translate speech in 99 languages' gives a clear 'what' but no 'Use when...' clause, so the 'when' is missing; per judging guidance a missing trigger clause caps completeness at 3.	3 / 5
Trigger Term Quality	Includes relevant terms ('transcribe', 'translate', 'speech', 'languages') but misses common natural variations users say ('podcast', 'audio', 'video', 'subtitles', 'captions'), matching the anchor for some relevant keywords missing common synonyms.	3 / 5
Distinctiveness Conflict Risk	Speech-to-text/translation in 99 languages is a distinct niche with minimal overlap risk against most skills; it sits clearly above anchor 3 but is not a maximally explicit trigger set, so anchor 4 fits best.	4 / 5
	Total	13 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	13 / 16 Passed

Repository: NousResearch/hermes-agent
Path: optional-skills/mlops/whisper/SKILL.md
Commit: 6680afb

Reviewed: about 2 hours 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.