Complete AI coding agent toolkit with skills, workflows, and agents for Claude Code
You are a visionary UI designer who creates interfaces that are not just beautiful, but implementable within rapid development cycles. Your expertise spans modern design trends, platform-specific guidelines, component architecture, and the delicate balance between innovation and usability. You understand that in the studio's 6-day sprints, design must be both inspiring and practical.
MUST BE USED for Distinguished Engineer level technical critiques, architecture reviews, technology assessments, and cost analysis. Use PROACTIVELY when evaluating major technical decisions, system design choices, technology stack selection, or preventing costly mistakes. Expert at challenging decisions with 25+ years perspective.
Universal API designer specializing in RESTful design, GraphQL schemas, and modern contract standards. **MUST BE USED** proactively whenever a project needs a new or revised API contract. Produces clear resource models, OpenAPI/GraphQL specs, and guidance on auth, versioning, pagination, and error formats—without prescribing any specific backend technology.
Elite systems architect ensuring clean, scalable designs that evolve gracefully. Expert in CQRS/ES, DDD, and microservices patterns. Use PROACTIVELY when creating services, modifying communication patterns, or changing architectural boundaries.
MUST BE USED to explore and document unfamiliar, legacy, or complex codebases. Use PROACTIVELY before refactors, onboarding, audits, or risk reviews. Produces a full-length report—architecture, metrics, risks, and a prioritised action plan—that other sub-agents can act on.
Set up or disable Langfuse observability for Claude Code sessions. Manages hook configuration, credential verification, and connection testing.
iOS 26 Liquid Glass design system for SwiftUI. Implements Apple's glassmorphism material effects, depth-based layering, and adaptive tinting. Follows Apple Human Interface Guidelines for glass materials. Use when: (1) Building iOS 26+ SwiftUI interfaces, (2) Implementing glassmorphism effects, (3) Creating translucent/frosted UI elements, (4) Designing with Apple's Liquid Glass aesthetic, (5) User mentions liquid glass, glassmorphism, or frosted glass UI.
List All Agent Worktrees
Produce a fully-filled autonomous-run mega-prompt at .agents/goals/<slug>.md by interviewing the user for the missing context. The output file is the "set the goal · walk away · come back to shipped work" prompt — drop it into a fresh Claude Code / Codex session to run end-to-end without hand- holding. Use when Stevie says "/make-a-goal", "/make-a-goal <outcome>", "make a goal", "set a goal", "build me a goal prompt", or similar. The skill does NOT execute the goal — it only produces the artifact. Named `make-a-goal` (not `goal`) because `/goal` is a reserved system slash.
Review codebase and create GitHub issues for identified problems
Uses power tools
Uses Bash, Write, or Edit tools
Has parse errors
Some configuration could not be fully parsed
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
One canonical source. Many AI tools. Write a skill once, install it into Claude Code, Codex, Copilot, Gemini, Hermes, nanoclaw, Amazon Q, Cursor, Cline, Roo, or Clawdhub.
This repository is the sole canonical home for the skills, the legacy
bootstrap.js installer, and the catalog. It is consumed as an external
source by ainb
(the Rust unit manager) and is installable directly by any SKILL.md-aware
agent harness.
| Repo | What it holds |
|---|---|
| stevengonsalvez/ainb-toolkit (this repo) | Canonical skills (skills/), agents (agents/), workflows, utilities, the external-dependencies.yaml manifest, the bootstrap.js legacy installer, and the generated catalog.yaml. |
| stevengonsalvez/agents-in-a-box | The ainb TUI/CLI unit manager (Rust). Installs/browses skills from this repo as a pinned external source, generates the enriched catalog-index.json release asset, and ships the reflect plugin + reflect-kb retrieval CLI. |
ainb (recommended)ainb (from the agents-in-a-box repo) treats this repo as an external
skill provider and handles install / update / sync / drift across tool
home dirs:
# Browse this repo's curated catalog
ainb skill browse "" --catalog ainb
# Add this repo as a source, then install any skill
ainb source add gh:stevengonsalvez/ainb-toolkit
ainb skill install gh:stevengonsalvez/ainb-toolkit@main/skills/<name> --targets claude --yes
# Verify + sync changes back to this repo
ainb skill check
ainb skill sync --to-repo
See the agents-in-a-box repo for ainb install (homebrew tap or
cargo install).
The flattened skills/ and agents/ trees are exposed as Nix packages:
inputs.ainb-toolkit.url = "github:stevengonsalvez/ainb-toolkit";
# ...
skillsPkg = ainb-toolkit.packages.${system}.skills; # $out/skills/<name>/...
agentsPkg = ainb-toolkit.packages.${system}.agents; # $out/agents/<cat>/...
nix build github:stevengonsalvez/ainb-toolkit#skills
bootstrap.js)The node installer still works and deploys the full tree to a tool home:
npm install
# Interactive (picks tool from menu)
node bootstrap.js
# Non-interactive (one tool)
node bootstrap.js --tool=claude-code-4.5
node bootstrap.js --tool=codex
node bootstrap.js --tool=copilot
node bootstrap.js --tool=gemini
# Verify everything landed (read-only)
node bootstrap.js --tool=claude-code-4.5 --verify
# Install external deps (git-cloned skills, npx-skills, claude plugins)
bash ~/.claude/setup-external.sh
Repeat for each target tool. Each tool gets its own home directory
population (~/.claude/, ~/.codex/, …). Hermes is special — it reads
from ~/.claude/skills/ via skills.external_dirs, so installing for
Claude transitively covers Hermes.
This repo is flattened: the canonical source trees live directly at
the repo root (there is no packages/ wrapper). Everything here gets
distributed to tool home dirs on bootstrap (or installed individually
by ainb).
skills/ 91 skills — agent-invokable SKILL.md bundles
agents/ 37 agents across 6 categories
├── design/ UI/UX designers
├── engineering/ backend, frontend, security, perf, code-review
├── meta/ agentmaker (create new agents)
├── orchestrators/ tech-lead, project-analyst, team-configurator
├── swarm/ worker/leader primitives for multi-agent work
└── universal/ backend-developer, frontend-developer, superstar-engineer
utilities/
├── config/ Shared tool config templates
├── hooks/ Event hooks (session-start, pre-commit, etc.)
├── output-styles/ Custom output formatters
├── reflections/ Session reflection templates
└── utils/ Shell utility libraries
workflows/
├── single-agent/ Guided plan → implement → validate flows
└── multi-agent/ Swarm/DAG orchestration commands
knowledge/
└── docs-solutions-template/
Per-tool config trees (claude-code-4.5/, codex/, copilot/,
gemini/, amazonq/, cursor/, cline/, roo/, clawdhub/, …) and
the shared general-rules/ also live at the repo root.
The reflect Claude Code plugin and the reflect-kb retrieval CLI are NOT in this repo — they live in the agents-in-a-box monorepo (
plugins/reflect/andreflect-kb/).bootstrap.jsinstalls reflect-kb viauv tool installby cloning that monorepo; see theinstallReflectKb()function for details.
npx claudepluginhub stevengonsalvez/ainb-toolkitBrowser automation and inspection toolkit for AI agents. Debug web apps, inspect UI, capture network traffic, monitor navigation, take screenshots.
Agent self-improvement + retrieval. Captures learnings via colon-namespaced sub-skills (reflect, reflect:consolidate, reflect:ingest, reflect:recall, reflect-status) and auto-injects relevant prior learnings via SessionStart/UserPromptSubmit hooks. PostToolUse arms low-cost mini-learning capture; Stop enqueues short-session reflection. v4 cost rearchitecture: a $0 skip-gate + cascade slice the drain so it runs on Sonnet under hard caps (8 turns / 180s / 2M-token poison), with cost observability (reflect cost) and a weekly Opus synthesis pass. Philosophy: Correct once, never again; recall everything next time.
Agent self-improvement + retrieval. Captures learnings via colon-namespaced sub-skills (reflect, reflect:consolidate, reflect:ingest, reflect:recall, reflect-status) and auto-injects relevant prior learnings via SessionStart/UserPromptSubmit hooks. PostToolUse arms low-cost mini-learning capture; Stop enqueues short-session reflection. v4 cost rearchitecture: a $0 skip-gate + cascade slice the drain so it runs on Sonnet under hard caps (8 turns / 180s / 2M-token poison), with cost observability (reflect cost) and a weekly Opus synthesis pass. Philosophy: Correct once, never again; recall everything next time.
Harness-native ECC plugin for engineering teams - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Plugin-safe Claude Code distribution of Antigravity Awesome Skills with 1,642 supported skills.