ai-engineer

Build production-ready LLM applications, advanced RAG systems, and intelligent agents. Implements vector search, multimodal AI, agent orchestration, and enterprise AI integrations.

Quality

38%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/antigravity-ai-engineer/SKILL.md

Quality

Content

10%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 reads as a persona/resume document rather than an operational skill: it is dominated by catalogs of models, frameworks, and traits Claude already knows, with zero executable guidance. Workflow steps exist only at an abstract level with no validation checkpoints, and nothing is split into reference files.

Suggestions

Delete or drastically compress the Capabilities, Knowledge Base, and Behavioral Traits sections — Claude already knows these technologies — and keep only the skill-specific workflow, conventions, and non-obvious decisions.

Replace the abstract Instructions with a concrete, sequenced workflow (e.g., requirements → architecture → implementation → validation) that includes explicit validation/verification checkpoints and executable commands or code patterns for the common cases (e.g., a reference RAG pipeline snippet, a guardrail/config checklist).

Move domain detail (model matrix, vector-DB comparison, RAG pattern catalog) into one-level-deep reference files (e.g., references/models.md, references/rag-patterns.md) linked from a lean SKILL.md overview, and remove time-sensitive model/version listings or isolate them in an explicitly dated section.

Dimension	Reasoning	Score
Conciseness	Roughly 110 of ~180 lines enumerate knowledge Claude already has ("OpenAI GPT-4o/4o-mini, o1-preview", "Vector databases: Pinecone, Qdrant, Weaviate", "Similarity metrics: cosine, dot product, Euclidean") plus redundant "Knowledge Base" and "Behavioral Traits" padding, and time-sensitive model versions ("Claude 4.5 Sonnet") appear outside any deprecated/old-patterns section — severely verbose, matching the lowest anchor.	1 / 5
Actionability	There is no code, command, or concrete example anywhere in the body; the Instructions ("Clarify use cases... Design the AI architecture... Implement with monitoring") only describe at a high level rather than instruct, matching the entirely-vague anchor.	1 / 5
Workflow Clarity	Two rough sequences exist (Instructions and the 8-step Response Approach) but steps are abstract and poorly defined, and validation checkpoints are absent despite production rollout and batch-oriented scope — matching the rough-sequence-many-gaps anchor.	2 / 5
Progressive Disclosure	The skill is a single monolithic file with no references/ or scripts/ bundle and no pointers to separate files, while the long Capabilities catalogs clearly belong in dedicated reference documents; section headers provide some structure, keeping it just above the no-structure anchor.	2 / 5
	Total	6 / 20 Passed

Description

66%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 communicates a clear capability set in third person with several concrete actions and decent trigger keywords, but it entirely lacks a "Use when..." clause, which both caps completeness and weakens its trigger function. Some phrases ("intelligent agents", "enterprise AI integrations") drift toward buzzword territory rather than concrete, searchable capabilities.

Suggestions

Add an explicit trigger clause, e.g. "Use when building LLM features, RAG pipelines, AI agents/chatbots, or vector search, or when the user mentions embeddings, retrieval, prompt engineering, or model APIs."

Replace buzzword phrasing ("intelligent agents", "enterprise AI integrations") with concrete techniques users would name, such as chunking, reranking, embeddings, function calling, or model routing.

Include a few natural synonyms users actually say ("chatbot", "GPT/Claude API", "retrieval-augmented generation") to broaden trigger-term coverage.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ("Build production-ready LLM applications", "Implements vector search, multimodal AI, agent orchestration") matching the several-specific-actions anchor, though "enterprise AI integrations" is buzzword-adjacent and no concrete techniques (chunking, reranking, embeddings) are named, keeping it below comprehensive.	4 / 5
Completeness	The "what" is clearly stated (builds LLM applications, RAG systems, agents, implements vector search), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Good natural keyword coverage ("LLM applications", "RAG", "agents", "vector search") that users would plausibly say, but common variations like "chatbot", "embeddings", "prompt", or "OpenAI/Claude API" are missing.	4 / 5
Distinctiveness Conflict Risk	The LLM-app/RAG/agent niche has distinct triggers and mostly minor overlap risk with related ML or data-science skills, though the broad "enterprise AI integrations" phrase invites some overlap.	4 / 5
	Total	15 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: boisenoise/skills-collections
Path: skills/antigravity-ai-engineer/SKILL.md
Commit: 7390864

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

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.