CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

whisper

Transcribe audio files to text using OpenAI Whisper

Quality

75%

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

Fix and improve this skill with Tessl

tessl review fix ./examples/skill/skills/whisper/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

85%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 strong, highly actionable reference: executable commands that match the bundled script, a useful model-size table, and clean section organization with an appropriately placed one-level script reference. The only real weakness is embedded time-sensitive version/date information and minor duplication of the frontmatter description.

Suggestions

Remove the 'as of 2026-11' date comment or relocate version-pin guidance to a dedicated section so the pinned version note doesn't silently rot.

Delete the opening line that restates the frontmatter description, and trim the Capabilities list that duplicates the Usage/Parameters sections.

DimensionReasoningScore

Conciseness

The body is mostly lean (commands, parameters, a model table, no explanations of concepts Claude already knows), but it embeds time-sensitive information — "openai-whisper 20250625 (pinned; latest release on PyPI as of 2026-11)" — outside any deprecated/old-patterns section, which the guidelines explicitly penalize, and the opening line "Transcribe audio files to text using OpenAI Whisper" duplicates the frontmatter description. Fits 'Mostly efficient but includes some unnecessary explanation or could be tightened'. Not 4 because the version/date staleness penalty applies; not 2 because there is no padded conceptual explanation.

3 / 5

Actionability

Fully executable, copy-paste-ready commands are given ("python3 scripts/transcribe.py audio.mp3 transcript.txt --model medium", plus --language, --timestamps, and --format json examples) covering the common cases, and the documented CLI exactly matches the bundled scripts/transcribe.py (verified: positional audio_file/output_file, --model choices, --language defaulting to auto, --timestamps, --format text/json). Matches 'Fully executable; copy-paste ready code or commands; specific examples cover the common cases'. Not 4 because there are no gaps — even installation commands are provided.

5 / 5

Workflow Clarity

This is a simple single-task skill (run one script), and the single action is unambiguous with documented parameters and a verified script path, so the simple-skill exception applies. The operation is neither destructive nor batch, so no validation cap applies. Not 4 because there is no sequencing ambiguity to leave gaps in.

5 / 5

Progressive Disclosure

Scored against the actual bundle: the body references scripts/transcribe.py, which exists and matches the documented interface, and the reference is one level deep and clearly signaled in every usage example. Sections (Usage, Parameters, Model Sizes, Dependencies, Installation) are well organized with no content that belongs in a separate file. Matches 'Clear overview with well-signaled one-level-deep references; content appropriately split'. Not 4 because no organization gaps were found.

5 / 5

Total

18

/

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, concise, and names a concrete action with a specific tool, but it is a single-clause statement with no 'when to use' guidance and no synonym/extension coverage. It reads as a bare capability label rather than a trigger-optimized description.

Suggestions

Add a 'Use when...' clause with explicit triggers, e.g. 'Use when the user mentions transcription, speech-to-text, converting audio to text, or .mp3/.wav files'.

Include natural synonyms and file extensions users would actually say (speech-to-text, voice memos, subtitles, .mp3, .wav, .m4a) to broaden trigger coverage.

Mention one or two additional concrete capabilities (timestamp/segment output, JSON format, language auto-detection) to raise specificity beyond a single action.

DimensionReasoningScore

Specificity

"Transcribe audio files to text using OpenAI Whisper" names the domain (audio files) and one concrete action (transcribe to text) plus the tool, matching the anchor 'Names domain and 1-2 concrete actions, but not comprehensive'. Not 4 because only a single action is listed — capabilities like timestamp generation, language selection, or output formats are absent from the description.

3 / 5

Completeness

The 'what' is clear ("Transcribe audio files to text using OpenAI Whisper"), but there is no 'Use when...' clause or equivalent trigger guidance, which the guidelines cap at 3 — an exact match for 'Has a clear what but when is missing or only weakly implied'. Not 4 because 'when' is entirely absent, not merely imprecise.

3 / 5

Trigger Term Quality

"Transcribe" and "audio files" are natural terms users would say, but the description misses common variations and synonyms such as "speech-to-text", "voice memo", "subtitles", or file extensions like .mp3/.wav, fitting 'Some relevant keywords but missing common variations or synonyms'. Not 4 because coverage is essentially a single phrasing rather than the multi-area keyword coverage of the 4 anchor.

3 / 5

Distinctiveness Conflict Risk

Naming the specific task (audio transcription) and tool ("OpenAI Whisper") makes this mostly distinct with only minor overlap risk against closely related speech-to-text skills, matching 'Mostly distinct; minor overlap risk'. Not 5 because it lacks the explicit distinct trigger phrases of the 5 anchor; not 3 because it is far more specific than generic examples like 'Works with document files'.

4 / 5

Total

13

/

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

trpc-group/trpc-agent-go

Path

examples/skill/skills/whisper/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.