- Home
- Skills
- Yanko Belov
- Code Craft
- Skill Awareness
skill-awareness_skill
- TypeScript
6
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill yanko-belov/code-craft --skill skill-awareness- SKILL.md4.4 KB
Overview
This skill tracks every application of code-craft skills across sessions to build a persistent practice record. It logs each skill use with timestamp, agent, project path, and a short context to enable analytics and surface underused skills. The goal is measurable improvement and clear auditing of applied practices.
How this skill works
When any code-craft skill is applied, the skill appends a single JSON line to a local JSONL file (~/.claude/skills/skill-usage.jsonl). Each entry records an ISO 8601 timestamp, kebab-case skill name, agent identifier, project path, and a brief context (<100 chars). Stored logs are then queryable with simple jq/bash commands for counts, recent entries, per-project breakdowns, and underused or never-used skill detection.
When to use it
- Always — log every explicit application of a code-craft skill
- After refactors that apply SOLID or core principles
- When adding or changing APIs, security, performance, or concurrency code
- When writing or modifying tests or error handling
- Before completing any coding task to verify checklist items
Best practices
- Follow the Iron Rule: never apply a skill silently — always log it
- Keep ctx concise and descriptive (under 100 characters) to preserve utility
- Use kebab-case for skill names to ensure consistent grouping
- Run provided jq commands to review totals, recent entries, and underused skills regularly
- If unsure whether a skill applies, prefer logging with a note of uncertainty
Example use cases
- Split a large class into two — log single-responsibility with a short context
- Add input validation to an endpoint — log input-validation and error-responses
- Introduce caching to prevent N+1 queries — log n-plus-one-prevention and caching
- Write unit tests using TDD — log tdd and test-isolation with project path
- Refactor authentication flows — log auth-patterns and secrets-handling
FAQ
Logs are appended as JSON lines to ~/.claude/skills/skill-usage.jsonl on the local machine.
What fields must each log entry include?
Each entry needs ts (ISO 8601), skill (kebab-case), agent (claude-code or opencode), project (path), and ctx (brief description).