Transform audio recordings into professional Markdown documentation with intelligent summaries using LLM integration
29
23%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/audio-transcriber/SKILL.mdQuality
Discovery
32%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 conveys the general purpose of converting audio to Markdown documentation but lacks the specificity and explicit trigger guidance needed for reliable skill selection. It uses somewhat buzzwordy language ('intelligent summaries', 'professional') without concrete actions, and critically omits a 'Use when...' clause that would help Claude know when to select this skill.
Suggestions
Add an explicit 'Use when...' clause with trigger scenarios, e.g., 'Use when the user wants to transcribe audio files, convert recordings to text, or generate meeting notes from audio.'
Include natural user-facing trigger terms like 'transcribe', 'transcription', 'meeting notes', 'voice recording', '.mp3', '.wav', 'speech-to-text'.
Replace vague qualifiers ('professional', 'intelligent') with specific concrete actions, e.g., 'Transcribes audio files to text, generates structured Markdown with headings and timestamps, and produces concise summaries of key topics.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (audio recordings, Markdown documentation) and mentions some actions (transform, summaries, LLM integration), but lacks specific concrete actions like 'transcribe speech', 'generate timestamps', 'extract key points', etc. | 2 / 3 |
Completeness | Describes what the skill does (transform audio to Markdown with summaries) but completely lacks a 'Use when...' clause or any explicit trigger guidance for when Claude should select this skill. Per rubric guidelines, missing 'Use when' caps completeness at 2, and the 'what' is also only moderately clear, so this scores a 1. | 1 / 3 |
Trigger Term Quality | Includes some relevant terms like 'audio recordings', 'Markdown documentation', and 'summaries', but misses common user-facing terms like 'transcribe', 'transcription', 'meeting notes', 'voice memo', '.mp3', '.wav', or 'speech-to-text'. | 2 / 3 |
Distinctiveness Conflict Risk | The combination of audio-to-Markdown is somewhat distinctive, but 'professional Markdown documentation' and 'intelligent summaries' are broad enough to overlap with general documentation or summarization skills. The audio focus helps but isn't sharply delineated. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
14%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is excessively verbose and poorly organized, with missing workflow steps (the actual transcription—Step 2—is absent), inconsistent numbering, mixed languages, and placeholder functions that aren't executable. While it demonstrates ambition in scope (metadata extraction, diarization, LLM summarization, batch processing), the content fails to deliver clear, actionable guidance and wastes significant token budget on terminal output mockups and installation flows that could be in separate files.
Suggestions
Fix the workflow numbering and add the missing Step 2 (actual transcription with Faster-Whisper/Whisper), which is the core operation of the skill.
Reduce content by 60-70%: remove terminal output mockups, consolidate examples to 1-2, move installation instructions to a separate INSTALL.md, and move the output template to a TEMPLATE.md.
Replace placeholder functions (cluster_by_topic, call_ai_model, extract_action_items) with actual executable implementations or remove them and provide a concrete, working approach.
Add explicit validation checkpoints, especially for batch processing (what happens if one file fails?) and after transcription completes (verify output file exists and is non-empty).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~400+ lines. Explains obvious concepts (what audio formats are, how file validation works), includes extensive UI mockups for terminal output, mixes languages (Portuguese/English), and repeats information. Much of this content (installation flows, progress bar displays, batch output examples) is unnecessary for Claude, who already knows how to structure CLI output and handle files. | 1 / 3 |
Actionability | Contains concrete bash and Python code snippets for detection, validation, and processing, but critical pieces are incomplete—`cluster_by_topic()`, `extract_action_items()`, `extract_decisions()`, and `call_ai_model()` are undefined placeholders. The workflow skips from Step 1 to Step 3 (missing Step 2 for actual transcription), and the LLM processing section references tools inconsistently. The code is a mix of executable and pseudocode. | 2 / 3 |
Workflow Clarity | Steps are numbered inconsistently (Step 0, Step 1, Step 3, Step 5—missing Steps 2 and 4), making the workflow confusing. The actual transcription step (the core operation) appears to be missing entirely. Scenario A for custom prompts appears mid-flow without clear sequencing. Batch processing is shown in examples but has no validation or error recovery steps, and there are no explicit checkpoints for the destructive/batch operations shown. | 1 / 3 |
Progressive Disclosure | Monolithic wall of text with no references to external files despite the content being long enough to warrant splitting. Installation instructions, output templates, LLM processing details, and multiple examples are all inline. No bundle files are provided or referenced, and the content would benefit greatly from separating the installation guide, output templates, and examples into separate files. | 1 / 3 |
Total | 5 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (560 lines); consider splitting into references/ and linking | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
76aea27
Table of Contents
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.