andrej-karpathy
Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.
64
76%
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 ./plugins/antigravity-awesome-skills-claude/skills/andrej-karpathy/SKILL.mdQuality
Discovery
67%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 is reasonably well-structured with a clear 'Use when...' clause and identifies its purpose as behavioral guidelines for LLM coding. However, the capabilities listed are more like abstract principles than concrete actions, and the broad trigger scope ('writing, reviewing, or refactoring code') could cause conflicts with other coding-related skills. The niche of 'reducing common LLM coding mistakes' is somewhat distinctive but could be articulated more sharply.
Suggestions
Add more specific, concrete actions the skill performs, e.g., 'Enforces minimal-change edits, flags unnecessary abstractions, requires explicit assumptions documentation, and generates testable acceptance criteria.'
Narrow the trigger terms or add distinguishing context to reduce conflict risk, e.g., 'Use when Claude is generating or modifying code and needs guardrails against common AI coding pitfalls like over-engineering or silent assumptions.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description names the domain (coding) and lists some actions ('writing, reviewing, or refactoring code') along with goals ('avoid overcomplication, make surgical changes, surface assumptions, define verifiable success criteria'), but these are more like behavioral principles than concrete, discrete actions. It's between vague and fully specific. | 2 / 3 |
Completeness | The description clearly answers both 'what' (behavioral guidelines to reduce common LLM coding mistakes) and 'when' (Use when writing, reviewing, or refactoring code), with explicit trigger guidance via the 'Use when...' clause. | 3 / 3 |
Trigger Term Quality | Includes relevant terms like 'writing', 'reviewing', 'refactoring', and 'code', which users might naturally say. However, it lacks common variations like 'debugging', 'programming', 'coding best practices', 'code quality', or 'code review'. The terms 'surgical changes' and 'surface assumptions' are less natural user language. | 2 / 3 |
Distinctiveness Conflict Risk | The description is about coding guidelines broadly, which could overlap with many other coding-related skills (e.g., code review skills, refactoring skills, code generation skills). The focus on 'LLM coding mistakes' and behavioral guidelines provides some distinction, but 'writing, reviewing, or refactoring code' is a very broad trigger that could conflict with more specific coding skills. | 2 / 3 |
Total | 9 / 12 Passed |
Implementation
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-crafted behavioral guideline skill that is concise, well-structured, and provides clear sequencing with verification steps. Its main weakness is that some guidance leans slightly abstract (e.g., the 'senior engineer' heuristic), and the examples, while helpful, could include more concrete before/after code snippets to maximize actionability. Overall, it's a strong skill that efficiently communicates its purpose.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is lean and efficient. It doesn't explain basic concepts Claude already knows. Each section delivers actionable behavioral rules without padding. The tradeoff note at the top is a single line. The examples are compact and illustrative. | 3 / 3 |
Actionability | The guidelines are concrete behavioral instructions with clear examples, but they are instruction-only (no executable code), which is appropriate for a behavioral skill. However, some guidance remains somewhat abstract—e.g., 'Ask yourself: Would a senior engineer say this is overcomplicated?' is subjective rather than mechanically actionable. The examples show input/output patterns but could be more specific with before/after code. | 2 / 3 |
Workflow Clarity | The skill presents a clear sequence (Think → Simplify → Surgical Changes → Goal-Driven Execution) with explicit verification checkpoints in section 4. The plan template with verify steps is a concrete feedback loop. For a behavioral guideline skill (not a destructive/batch operation), this level of workflow clarity is appropriate and well-structured. | 3 / 3 |
Progressive Disclosure | For a standalone skill under 80 lines with no need for external references, the content is well-organized into clearly labeled sections with logical progression. The structure is easy to scan with headers, bold callouts, and bullet points. No bundle files are needed or referenced, which is appropriate. | 3 / 3 |
Total | 11 / 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
- sickn33/antigravity-awesome-skills
- Commit
22710e9
- 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.