azure-ai-document-intelligence-ts
Extract text, tables, and structured data from documents using prebuilt and custom models.
55
62%
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/antigravity-azure-ai-document-intelligence-ts/SKILL.mdQuality
Discovery
60%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 does a good job listing specific extraction capabilities (text, tables, structured data) and hints at the mechanism (prebuilt and custom models). However, it lacks an explicit 'Use when...' clause which caps completeness, and the term 'documents' is too broad without specifying file types or common user scenarios. Adding trigger terms and explicit usage guidance would significantly improve skill selection accuracy.
Suggestions
Add a 'Use when...' clause with explicit triggers, e.g., 'Use when the user needs to extract information from documents, invoices, receipts, or forms using AI models.'
Include natural trigger terms users would say, such as specific document types (PDF, invoices, receipts, forms), file extensions (.pdf, .png, .jpg), and related terms like 'OCR', 'document parsing', 'form recognition'.
Clarify what 'prebuilt and custom models' refers to (e.g., Azure Document Intelligence, AWS Textract) to improve distinctiveness from generic document extraction skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'Extract text, tables, and structured data' and mentions both 'prebuilt and custom models' as methods. This provides clear, actionable capabilities. | 3 / 3 |
Completeness | Clearly answers 'what does this do' (extract text, tables, structured data from documents using models), but lacks an explicit 'Use when...' clause or trigger guidance for when Claude should select this skill. | 2 / 3 |
Trigger Term Quality | Includes some relevant keywords like 'extract text', 'tables', 'structured data', and 'documents', but misses common user variations like specific file types (PDF, invoices, receipts), OCR, or document parsing. The term 'prebuilt and custom models' is more technical jargon than natural user language. | 2 / 3 |
Distinctiveness Conflict Risk | The mention of 'prebuilt and custom models' adds some distinctiveness suggesting a specific document intelligence/AI extraction tool, but 'documents' is very broad and could overlap with PDF skills, OCR skills, or general data extraction skills. | 2 / 3 |
Total | 9 / 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, highly actionable SDK reference with excellent executable code examples covering the full breadth of Azure Document Intelligence features. Its main weaknesses are repetitive patterns across examples (the polling + error check pattern is shown ~8 times nearly identically), lack of validation/verification steps for results, and a monolithic structure that could benefit from splitting into overview + detailed reference files.
Suggestions
Reduce repetition by showing the full polling pattern once, then using abbreviated versions (e.g., '// ... poll as shown above') in subsequent examples to save significant token budget.
Add result validation guidance — e.g., checking confidence scores against thresholds, verifying expected field presence, and handling partial/empty results.
Split into SKILL.md (overview + quick start + model table) and a REFERENCE.md (detailed examples for each model type, custom models, classifiers) to improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient with executable code examples, but there's significant repetition in the polling pattern (shown in nearly every example then again as a standalone section), and the invoice/receipt extraction examples are very similar to the general analyze examples. The 'Best Practices' and 'Limitations' sections contain some generic advice Claude already knows. | 2 / 3 |
Actionability | Every section provides fully executable, copy-paste ready TypeScript code with proper imports, type annotations, and error handling. The examples cover authentication, URL/local file analysis, prebuilt models, custom models, classifiers, and pagination — all with concrete, runnable code. | 3 / 3 |
Workflow Clarity | The polling pattern section clearly sequences the async workflow (start → check errors → create poller → monitor → wait), but there are no validation checkpoints for verifying results quality, handling partial failures, or retry logic for transient errors. The custom model build workflow lacks validation of training data or model quality assessment steps. | 2 / 3 |
Progressive Disclosure | The content is well-organized with clear section headers and a useful prebuilt models reference table, but at ~250 lines it's a monolithic file with no references to supporting documents. The invoice and receipt extraction examples could be split into a separate examples file, and the custom model/classifier sections could be in an advanced guide. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- boisenoise/skills-collections
- Commit
8ac11ab
- 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.