Claude Skills
The open format is called Agent Skills and works in Claude Code, Codex, Cursor and other agents — most people know it as Claude Skills.
Every Agent Skill we could find on GitHub, deduplicated by content. 79 600 files from 1 763 authors, of which 61 947 are unique — the rest is the same skill repackaged into someone else's repository. For each one: what it weighs in tokens, whether it ships runnable scripts, and which MCP servers it needs.
Kubernetes operations: debugging, security, RBAC, and infrastructure tooling.
CPU-only motion data processing pipeline for game animation: BVH import, contact detection, root decomposition, motion blending, FABRIK IK. No GPU required.
Phaser 3 2D game dev: scenes, physics, tilemaps, sprites, polish.
Audit and create cron jobs with reliability and safety.
Audit CVE/vulnerability source coverage for a technology stack. Maps each component (container, library, base image, runtime) to authoritative CVE feeds, flags gaps, and produces audit-ready reports. Generic: works for any service or stack.
Deterministic API endpoint validation with pass/fail reporting.
Generate headless Claude Code cron jobs with safety.
Publish a public website safely: DNS, web server, HTTPS, hardening, verify. Routes raw dev servers through nginx/Caddy/Apache/Cloudflare Pages.
Service health monitoring, endpoint validation, and CVE source auditing.
Shell configuration: Fish and Zsh setup, PATH, completions, plugins.
Evaluate agents and skills for quality and standards compliance.
Safely start, supervise, and terminate shell processes: background jobs, PID capture, signals, traps, cleanup verification.
A/B test agent variants for quality and token cost.
Scaffold vexjoy-agent operator .md files: frontmatter, routing block, operator context, reference loading table, phase/gate workflow.
Run benchmark-selected GPT-5.6 work through the Codex CLI.
Classify user requests and route to the correct agent + skill. Primary entry point for all delegated work.
Background memory consolidation and learning graduation — overnight knowledge lifecycle.
Analyze agent/skill reference depth and generate missing domain-specific reference files.
Detect documentation drift against filesystem state.
Maintain /do routing tables when skills or agents change.
Generate project-specific CLAUDE.md from repo analysis.
Query and display structured decision traces from routing, agent selection, and skill execution.
Verify VexJoy Agent installation, diagnose issues, and guide first-time setup.
| Generate rich self-contained HTML artifacts instead of markdown. Auto-detects artifact shape (spec, code-review, prototype, report, editor, data-viz, diagram, deck) and loads shape-specific patterns. Bundles Birchline design system with 4 theme presets. Use for "make HTML", "as HTML", "HTML artifact", or auto-injected by router when output benefits from rich visualization.
Manually teach error pattern and solution to learning database.
Loop /do cycles until done-criteria verify or budget stops.
Learning system interface: stats, search, graduate, clear learnings. Backed by learning.db (SQLite + FTS5).
|
DAG-based multi-skill orchestration with dependency resolution.
Create and iteratively improve skills through eval-driven validation.
Closed-loop toolkit self-improvement: discover gaps, diagnose, propose, critique, build, test, evolve.
Evaluate skills: trigger testing, A/B benchmarks, structure validation, head-to-head bake-offs.
Interactive guide to workflow system: agents, skills, routing, execution patterns.
Multi-agent consultation for architecture decisions.
Polling, retry, and backoff patterns.
Post-mortem diagnostic analysis of failed workflows.
Collaborative coding with enforced micro-steps and user-paced control.
Triage GitHub notifications and issue/PR queues.
Question-only debugging: guide users to find root causes themselves.
Capture forward-looking idea as a seed for future feature design.
Planning lifecycle: specs, requirements, pre-plan ambiguity resolution, file-backed plans, plan validation, pause/resume, session handoff.
Tracked lightweight execution with composable rigor flags: --trivial, --discuss, --research, --full. Covers zero-ceremony inline fixes (typo, spelling fix, small mistake in a single file, ≤3 edits) through contained multi-file changes.
| cleanup, and PR mining. Use when user wants to commit changes, get a second-opinion code review from Codex, push changes, create a PR, check PR status, fix review comments, clean up branches after merge, or mine tribal knowledge from PR reviews. Use for "commit my changes", "codex review", "push my changes", "create a PR", "pr status", "fix PR comments", "clean up branches", "mine PRs", or "address feedback".
Statistical rule discovery from Go codebase patterns.
Systematic codebase exploration and architecture mapping.
Read-only exploration, inspection, and reporting without modifications. Explore and report on code/config/state without writing or modifying anything.
Package session state for the next agent, or rehydrate it at start.
Mandatory rules for agents in git worktree isolation.
Anti-rationalization enforcement for maximum-rigor task execution.
Proactive architecture improvement: find shallow modules, propose deepening opportunities, design conversation.
Comprehensive four-wave review of all repo source files, producing a prioritized issue backlog.
Collect, filter, and freshness-qualify news items.
Decision-first data analysis with statistical rigor gates.
Fresh-subagent-per-task execution with two-stage review gates.
Defense-in-depth verification before declaring any task complete.
Weighted decision scoring for architectural choices.
Extract video transcripts: yt-dlp subtitles to clean paragraphs.
Verify cross-component wiring and data flow.
Verify factual claims against sources before publish.
Constructive critique via 5 HackerNews personas with claim validation.
Claude Skills — questions
Answers built from the skills we actually parsed.