Expert in building products that wrap AI APIs (OpenAI, Anthropic, etc. ) into focused tools people will pay for. Not just "ChatGPT but different" - products that solve specific problems with AI.
35
32%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/antigravity-ai-wrapper-product/SKILL.mdQuality
Discovery
22%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 reads more like a marketing tagline than a functional skill description. It lacks concrete actions, has no 'Use when...' clause, and relies on vague language like 'focused tools people will pay for' without specifying what the skill actually does (e.g., architecture design, prompt engineering, pricing strategy, API integration patterns).
Suggestions
Add a 'Use when...' clause with specific trigger scenarios, e.g., 'Use when the user wants to build a SaaS product on top of AI APIs, design an AI wrapper, or monetize AI capabilities.'
Replace vague language with concrete actions, e.g., 'Designs API integration architectures, creates prompt pipelines, structures pricing models, and builds user-facing AI features.'
Use third-person voice consistently and add natural trigger terms users would say, such as 'AI SaaS', 'API wrapper', 'AI product', 'LLM integration', 'AI startup MVP'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description uses vague language like 'building products' and 'focused tools' without listing concrete actions. It describes a general philosophy ('not just ChatGPT but different') rather than specific capabilities like 'designs API integration architectures, creates pricing models, builds prompt pipelines.' | 1 / 3 |
Completeness | The 'what' is vaguely stated (building products that wrap AI APIs) but lacks specifics, and there is no 'when' clause or explicit trigger guidance at all. The missing 'Use when...' clause caps this at 2 per the rubric, but the weak 'what' brings it to 1. | 1 / 3 |
Trigger Term Quality | It includes some relevant keywords like 'AI APIs', 'OpenAI', 'Anthropic', and 'products' that users might mention. However, it misses common variations like 'SaaS', 'wrapper', 'API integration', 'monetization', 'prompt engineering', or 'AI startup'. | 2 / 3 |
Distinctiveness Conflict Risk | The focus on wrapping AI APIs into paid products provides some distinctiveness, but 'building products' is broad enough to overlap with general software development, product management, or other AI-related skills. | 2 / 3 |
Total | 6 / 12 Passed |
Implementation
42%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill provides genuinely useful, executable code examples for building AI wrapper products, covering cost tracking, rate limiting, streaming, and output validation. However, it is severely bloated — repeating headers, explaining concepts Claude already knows (model comparisons, what thin wrappers are), and including meta-sections (Expertise, Capabilities, When to Use, Limitations) that add little instructional value. The monolithic structure with no bundle files or cross-references makes it hard to navigate and wastes context window budget.
Suggestions
Cut the file by 50%+: remove the Expertise/Capabilities lists, model comparison tables (Claude knows these), 'When to Use' triggers, 'Limitations' boilerplate, and deduplicate repeated section headers like 'AI Product Architecture'.
Split into bundle files: move Sharp Edges/cost management to COST-MANAGEMENT.md, rate limiting to RATE-LIMITS.md, and prompt engineering patterns to PROMPTS.md, with brief summaries and links in the main SKILL.md.
Add an explicit end-to-end workflow with validation checkpoints: e.g., '1. Design prompt template → 2. Test with sample inputs → 3. Validate output schema → 4. Add cost tracking → 5. Set usage limits → 6. Load test before launch'.
Remove or drastically shorten the Collaboration and Delegation Triggers sections — these are routing metadata, not instructional content, and consume significant tokens.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~400+ lines. Repeats section headers (e.g., 'AI Product Architecture' appears twice), explains obvious concepts like what thin wrappers are, includes lengthy tables of common knowledge (model comparisons Claude already knows), and has redundant sections like 'Capabilities' and 'Expertise' that overlap heavily. The collaboration/delegation section and 'When to Use' triggers add significant bulk with minimal instructional value. | 1 / 3 |
Actionability | Provides fully executable JavaScript code examples throughout — API calls with the Anthropic SDK, retry logic with exponential backoff, request queuing with p-queue, streaming responses, caching patterns, cost tracking, and usage limits. Code is copy-paste ready with concrete implementations rather than pseudocode. | 3 / 3 |
Workflow Clarity | The wrapper stack diagram provides a clear high-level sequence, and individual code blocks show specific steps. However, there's no unified workflow tying the pieces together with explicit validation checkpoints. The 'Sharp Edges' sections describe problems and fixes but don't integrate into a coherent build-validate-deploy workflow. For a skill involving API cost management and destructive billing consequences, the lack of explicit verification steps between stages caps this at 2. | 2 / 3 |
Progressive Disclosure | Everything is crammed into a single monolithic file with no references to external files despite the content being extensive enough to warrant splitting (prompt engineering, cost management, rate limiting, and differentiation could each be separate files). The collaboration section references other skills but no supporting bundle files exist. The document is a wall of text with no clear navigation hierarchy. | 1 / 3 |
Total | 7 / 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 (689 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 | |
e41e34a
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.