deepgram-core-workflow-a
Implement production pre-recorded speech-to-text with Deepgram. Use when building audio transcription, batch processing, or implementing diarization and intelligence features. Trigger: "deepgram transcription", "speech to text", "transcribe audio", "batch transcription", "deepgram nova", "diarize audio".
80
77%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./plugins/saas-packs/deepgram-pack/skills/deepgram-core-workflow-a/SKILL.mdQuality
Discovery
89%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a well-structured skill description that clearly identifies its niche (Deepgram pre-recorded speech-to-text), provides explicit 'Use when' guidance, and includes strong trigger terms covering both brand-specific and generic user language. The main weakness is that the capability listing could be more concrete—specifying exact actions like 'parse transcription results', 'handle callback URLs', or 'configure Nova model parameters' would strengthen specificity.
Suggestions
Add more concrete actions beyond high-level categories, e.g., 'configure Nova models, parse transcript JSON responses, handle audio file uploads, set up callback webhooks' to improve specificity.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (speech-to-text with Deepgram) and some actions (audio transcription, batch processing, diarization, intelligence features), but doesn't list multiple concrete actions in detail—e.g., it doesn't specify output formats, supported audio types, or specific API operations. | 2 / 3 |
Completeness | Clearly answers both 'what' (implement pre-recorded speech-to-text with Deepgram) and 'when' (explicit 'Use when' clause with triggers for audio transcription, batch processing, diarization) plus an explicit trigger list. | 3 / 3 |
Trigger Term Quality | Includes a well-curated set of natural trigger terms users would actually say: 'speech to text', 'transcribe audio', 'batch transcription', 'deepgram nova', 'diarize audio', and 'deepgram transcription'. These cover both brand-specific and generic variations. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive due to the Deepgram-specific focus and pre-recorded speech-to-text niche. The trigger terms like 'deepgram nova' and 'deepgram transcription' clearly separate it from generic audio or other STT provider skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, actionable skill with production-ready TypeScript code covering multiple Deepgram use cases. Its main strengths are concrete, executable examples and a useful error handling table. Weaknesses include missing validation/verification steps in the batch workflow and some verbosity that could be trimmed, particularly in inline comments and the options interface.
Suggestions
Add a validation checkpoint after batch processing (e.g., check confidence thresholds, verify non-empty transcripts) and a retry mechanism for transient failures to improve workflow clarity.
Trim inline comments that restate obvious information (e.g., '// Audio intelligence', '// Topic detection') and remove explanatory comments Claude doesn't need (e.g., '// Deepgram returns a request_id immediately, processes in background').
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient with executable code and useful tables, but includes some unnecessary commentary (e.g., explaining what callback does, inline comments restating obvious things). The options interface lists every parameter with comments that Claude would already understand. Could be tightened by ~20%. | 2 / 3 |
Actionability | Fully executable TypeScript code throughout, with concrete class implementations, specific API calls, real configuration options, and copy-paste ready examples. The error handling table provides specific solutions including an ffmpeg command. Keywords show exact syntax with weight ranges. | 3 / 3 |
Workflow Clarity | Steps are clearly sequenced (build class → extract results → batch → async → keyword boosting), but there are no validation checkpoints. Batch processing uses Promise.allSettled which is good, but there's no guidance on verifying transcription quality, no retry logic for transient failures, and no feedback loop for error recovery in the batch workflow. | 2 / 3 |
Progressive Disclosure | The content is well-structured with clear sections and a reference to a companion workflow (deepgram-core-workflow-b), plus external documentation links. However, the skill is quite long (~200 lines of code) and some sections like the full TranscribeOptions interface or the formatResult function could be split into referenced files. No bundle files exist to support progressive disclosure. | 2 / 3 |
Total | 9 / 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 |
|---|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Commit
3a2d27d
- 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.