catalinnita/coding-agents-skills
Skill Claude CodeCodex
Audits the codebase for WCAG 2.1 accessibility violations — missing alt text, broken keyboard navigation, colour contrast failures, invalid ARIA, and more — then applies safe targeted fixes and produces a severity-ranked HTML report. Use when asked to run an accessibility audit, fix accessibility issues, check WCAG…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Builds the project, measures every output asset (JS, CSS, images, fonts) in raw and gzip sizes, checks against configured thresholds, persists versioned JSON history, and produces a self-contained HTML report with an SVG treemap and size trend charts. Use when asked to analyse bundle sizes, check for size regressions…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
After any skill runs and modifies the codebase, generates a human-readable change log that explains every modification in plain language pitched at the configured developer level (absolute-beginner, junior, senior) with links to authoritative further reading for each change type. Use when asked to explain what a skill…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Records the token usage and USD cost of every Claude action (skill runs, agent phases, prompts) and persists them to a structured JSON file. Provides record, report, and reset modes. Use when asked to track Claude costs, record token usage, show Claude spend, summarise API costs, or how much an operation cost.
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scans a TypeScript/JavaScript repository for hardcoded values that belong in configuration — URLs, secrets, ports, timeouts, limits, feature flags, repeated literals, paths, and magic numbers — extracts them into .env.example, config/config.ts, config/feature-flags.ts, and config/constants.ts, then rewrites source…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Runs the test suite with coverage enabled, parses the results, identifies every file, function, and branch that is untested or below configured thresholds, and produces a prioritised structured gap report that the tdd-agent reads to create tests one by one. Use when asked to find coverage gaps, show untested code…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scaffold a new Agent Skill in the agentskills.io format. Use when asked to create a skill, add a skill, set up a skill folder, or build a new reusable agent capability. Generates a SKILL.md with correct YAML frontmatter, an optional scripts/ stub, and validates the result with skills-ref.
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scans all CSS/SCSS files in a repo, extracts hardcoded values (colors, spacing, typography, radii, shadows, z-index, transitions) into CSS custom properties, and replaces every occurrence with a variable reference. Supports format and unit normalization, perceptual color merging, grid and scale snapping to keep…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Converts any form of acceptance criteria or informal scenario descriptions into well-formed Gherkin-syntax scenario files, then validates them against the codebase to ensure full coverage including edge cases and error paths. Use when asked to generate cucumber scenarios, convert scenarios to Gherkin, write BDD…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Analyses an existing repository, scaffolds a Nextra documentation app in docs/, auto-generates MDX pages for getting-started, scripts, API, and components sections, and produces Excalidraw diagram files for data flows and state management. Use when asked to generate documentation, create docs for this repo, scaffold a…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scans React TSX/JSX files (or raw HTML) for structurally repetitive JSX subtrees, proposes reusable component extractions with inferred prop signatures, writes new component files, and rewrites the original files to use them. Use when asked to extract repeated JSX into components, reduce duplication in React files…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Identifies modules mocked individually in many test files and promotes them to global mocks via setupFiles or mocks directories so each test file gets them automatically. Removes redundant per-file mock declarations and wires up MSW server reset in afterEach. Use when asked to define global mocks, promote mocks to…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Reads the latest Lighthouse audit results, parses every failed or warning audit, applies auto-fixable recommendations directly to the source code (render-blocking scripts, missing preconnects, LCP preloads, font-display, image sizing, WebP conversion, text compression config, document.write replacement), and produces…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Runs Lighthouse CLI audits against configured URLs, averages results across multiple iterations, persists versioned JSON history, and produces a self-contained HTML comparison report covering Core Web Vitals, asset sizes, and score trends over time. Use when asked to run lighthouse tests, check core web vitals, check…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Measures server-side rendering time and origin vs CDN request counts for configured URLs using HTTP timing and Playwright network interception, persists versioned JSON history, and produces a self-contained HTML report with render time and request count trends. Use when asked to measure page render time, check SSR…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Audits the codebase for security vulnerabilities including OWASP Top 10 issues, committed secrets and credentials, vulnerable dependencies (CVEs), and missing HTTP security headers. Applies safe targeted fixes and produces a severity-ranked HTML report with remediation guidance. Use when asked to run a security audit…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Audits the codebase for SEO issues — missing or malformed metadata, poor heading structure, absent structured data, crawlability problems, and Core Web Vitals heuristics — then applies safe fixes and produces a priority-ranked HTML report. Use when asked to run an SEO audit, fix SEO issues, add meta or open graph…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Reads a target skill's SKILL.md and scripts, then generates exhaustive test cases with mocks covering happy paths, edge cases, boundary values, error conditions, and all documented gotchas. Use when asked to write tests for a skill, generate test cases for a skill, or create a test suite for an agent skill.
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scans every test file to detect the dominant testing pattern (framework, describe/it structure, hook style, naming convention, async style, assertion style), identifies deviating files, and normalises them so the entire test suite follows the same conventions. Writes a test-conventions.json consumed by the tdd-agent…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Analyses all TypeScript source files in a repo, applies configurable type hygiene rules (deduplication, redundancy removal, enum extraction, normalisation, cleanup, organisation), and rewrites files in place. Use when asked to clean up types, deduplicate types, generate enums, unify declarations, or remove redundant…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Scans the test suite for all mock definitions, identifies duplication and inconsistency, consolidates them into shared typed mock factory functions, and replaces inline mock literals with calls to those factories. Use when asked to create unified mocks, consolidate mocks, extract shared mock factories, remove…
catalinnita/coding-agents-skills
Skill Claude CodeCodex
Captures screenshots of the current branch and the main/master branch using git worktrees, compares them pixel-by-pixel and perceptually (SSIM), and produces a self-contained HTML diff report. Use when asked to run visual regression, check for visual regressions, compare screenshots with main, or visual diff a branch.
catalinnita/coding-agents-skills
MCP server Claude CodeCodexCursor
MCP server "vercel", hosted remotely at mcp.vercel.com, as configured in catalinnita/coding-agents-skills.