agentic-development
Build AI agents with Pydantic AI (Python) and Claude SDK (Node.js)
42
30%
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 ./skills/agentic-development/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 identifies a clear domain (AI agent building) and names specific frameworks, which is helpful for differentiation. However, it lacks concrete actions, natural trigger terms users would say, and critically has no 'Use when...' clause to guide skill selection. It reads more like a title than a functional description.
Suggestions
Add a 'Use when...' clause with explicit triggers, e.g., 'Use when the user asks about building AI agents, using Pydantic AI, the Anthropic Claude SDK, tool use, function calling, or agentic workflows.'
List specific concrete actions the skill covers, e.g., 'Define agent tools, configure agent loops, handle structured outputs, manage conversation state, implement streaming responses.'
Include common keyword variations users might naturally say, such as 'pydantic-ai', 'anthropic SDK', 'agentic', 'tool use', 'function calling', 'agent framework'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (AI agents) and specifies two frameworks (Pydantic AI, Claude SDK) with their languages, but doesn't list concrete actions like 'create tool definitions', 'configure agent loops', 'handle streaming responses', etc. | 2 / 3 |
Completeness | Describes what (build AI agents with specific frameworks) 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 thin, so this scores 1. | 1 / 3 |
Trigger Term Quality | Includes some relevant keywords like 'AI agents', 'Pydantic AI', 'Claude SDK', 'Python', 'Node.js', but misses common user terms like 'agentic', 'tool use', 'function calling', 'agent framework', 'pydantic-ai', or 'anthropic SDK'. | 2 / 3 |
Distinctiveness Conflict Risk | Mentioning specific frameworks (Pydantic AI, Claude SDK) provides some distinctiveness, but 'Build AI agents' is broad enough to overlap with other agent-building or SDK-related skills. The dual-framework scope also increases potential conflict with Python-specific or Node.js-specific skills. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
27%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill attempts to be a comprehensive guide to agentic development across multiple frameworks and languages, but suffers severely from trying to cover too much in a single file. The content is excessively verbose with many concepts Claude already understands (what agents are, what tools do, basic architecture patterns), and most code examples are illustrative pseudocode rather than executable implementations. The strongest sections are the concrete Pydantic AI and Claude SDK examples at the top, but these are buried in hundreds of lines of generic agent development advice.
Suggestions
Split content into separate files by framework (pydantic-ai.md, claude-sdk.md, gemini.md) and reference them from a lean SKILL.md overview, matching the 'Load with' pattern already mentioned in the header.
Remove explanatory content Claude already knows: agent architecture diagrams, what tools/memory/guardrails are conceptually, and the OpenAI 'Three Components' diagram. Focus only on implementation patterns.
Make code examples executable by either providing complete working implementations or explicitly noting which functions need user implementation. Replace pseudocode patterns like `llmCall()`, `createAgent()`, and `executeTool()` with real library calls.
Cut the model selection table and framework-agnostic advice (anti-patterns list, generic checklist) which are general knowledge, and focus the skill on the specific patterns and code needed to build agents with the two default frameworks.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~600+ lines. Includes extensive explanations of concepts Claude already knows (agent architecture diagrams, what tools/memory/guardrails are), covers multiple frameworks (Pydantic AI, Claude SDK, OpenAI, Gemini) with redundant patterns, and includes project structure templates that are generic knowledge. Much of this content could be cut by 60-70% without losing actionable value. | 1 / 3 |
Actionability | Contains many code examples that appear executable, but most are pseudocode-like patterns with undefined functions (executeTool, runCommand, llmCall, createAgent) and incomplete implementations. The Pydantic AI and Claude SDK examples at the top are the most concrete and copy-paste ready, but the bulk of the middle sections use illustrative TypeScript that wouldn't compile without significant additional code. | 2 / 3 |
Workflow Clarity | The Explore-Plan-Execute-Verify workflow is clearly sequenced with verification steps, which is good. However, the verification implementations are pseudocode with undefined helpers, and there are no concrete validation commands or checkpoints for the actual agent development process itself. The workflow pattern is described abstractly rather than being tied to specific, executable steps. | 2 / 3 |
Progressive Disclosure | This is a monolithic wall of text with everything inline. Despite the header mentioning 'Load with: base.md + llm-patterns.md + [language].md', the skill itself contains all content for multiple languages, multiple frameworks, and multiple concerns (architecture, tools, memory, guardrails, testing, model selection) that should be split into separate referenced files. No content is delegated to external files despite the massive length. | 1 / 3 |
Total | 6 / 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 (857 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 | |
- Repository
- alinaqi/claude-bootstrap
- Commit
d4ddb03
- 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.