Curated power-ups for coding agents: skills, slash commands, MCP configs, hooks, AGENTS.md templates, and workflows for serious software engineering. Claude Code, Codex, Antigravity CLI, Cursor and more
> /plugin marketplace add yeaight7/agent-powerups
Repo: yeaight7/agent-powerups
What's inside
Agent Powerups is an Oh My Zsh-style collection of reusable skills, slash commands, MCP configs, hooks, AGENTS.md templates, and workflows for coding agents.
Today, this repo ships:
apx) with runnable local checksapx plugins inspection (apx plugins list)apx profiles for curated skill/plugin setsNative install is direct for humans. Safety boundaries stay around external tools, secrets, shell profiles, and MCP enablement.
| Path | Status | Notes |
|---|---|---|
skills/ | shipped | Reusable agent workflows such as systematic-debugging and writing-plans |
mcp/ | shipped | Local-first GitHub MCP config with check, smoke, and explicit install commands |
agents-md/ | shipped | Starter AGENTS.md templates |
commands/ | shipped | Review-first command prompts plus safe runnable checks |
hooks/ | shipped | Review-before-use hook recipes plus safe runnable checks |
workflows/ | shipped | Scenario guides |
plugins/ | shipped | Plugins with local-first discovery, validation, native install, and marketplace metadata (apx plugins list) |
scripts/ | shipped | Validation and tool-check helpers for this repo |
examples/ | shipped | Minimal safe setup examples |
If you are an agent working in a repo with Agent Powerups available, route tasks to capabilities yourself instead of waiting to be told:
Match first. On Claude Code the host already lists skills natively — match the task to the narrowest fitting skill (plus obvious fits like no-fluff). On Codex, Gemini, or generic agents, start with apx discover.
Discover the rest — anything not natively loaded (MCP configs, hooks, AGENTS.md templates, installed-only assets):
apx discover "<the user's task>" --target <codex|claude-code|gemini|generic> --json
apx info <chosen-asset> # what it is, when to use it, source path, next action
Read, apply, verify. Read the asset's SKILL.md (or file) before applying it; run apx check <asset> only if it declares external requirements; then verify the result against the task.
Gate risky assets. MCP configs and hooks are review-before-use — surface them for approval, never auto-enable.
npm install -g agent-powerups
apx install claude --full
This installs the apx CLI globally and runs a full Claude Code setup — copies skills, plugins, and commands into ~/.claude/ and patches your CLAUDE.md. Use --dry-run to preview changes first.
git clone https://github.com/yeaight7/agent-powerups.git
cd agent-powerups
npm install
npm run build
npm link
apx doctor
apx doctor --full
apx discover "fix a failing test regression" --target codex
apx inventory --target codex --json
apx list
apx list --json --verbose
apx info markitdown-file-intake
apx commands run ship-check
apx hooks run no-secrets-preflight --all
apx mcp check github-local --target generic
apx mcp smoke github-local --json
apx mcp install github-local --target codex --dry-run
apx install codex --dry-run
apx install codex
apx install claude
apx install gemini
apx install codex --full
apx install codex --verbose
Default native install copies all root skills and plugins into the selected agent root and writes a discovery-index.json beside them. Human output shows counts by default; use --verbose for per-file paths. If installed Agent Powerups guidance or discovery indexes are stale, apx install <agent> refreshes them by default. --full also stages support assets and another discovery index under agent-powerups/, then updates existing global instructions with a backup.
apx plugins list
apx plugins info dev-vitals
apx plugins validate --all
apx plugins install dev-vitals --target codex --dry-run
apx ask-codex "Return OK only" --json
apx ask-claude "Return OK only" --json
apx ask-gemini "Return OK only" --json
apx relay init second-opinion
apx relay start second-opinion --provider gemini
apx relay ask second-opinion "Review this plan" --json
apx relay status second-opinion
apx relay stop second-opinion
apx profiles list
apx profiles info safe-core
apx profiles plan safe-core --target codex
apx check markitdown-file-intake
apx check graphify
Use apx check only for assets that declare external requirements. A successful dependency check does not mean the skill or workflow was used correctly.
Preview supported dependency installers before asking for approval:
apx check defuddle --install-missing --dry-run
apx check markitdown-file-intake --install-missing --dry-run
apx check graphify --install-missing --dry-run
apx install markitdown-file-intake --target codex --dry-run
apx install ask-claude --target codex --dry-run
apx setup codex --dry-run
apx setup codex --mode minimal --yes # bootstrap only
apx setup codex --mode recommended --yes # main agent setup (recommended)
apx setup codex --mode full --yes # broad staging
apx setup is legacy compatibility and remains supported through at least v0.8.0. Prefer apx install <agent> for manual native install.
apx install <codex|claude|claude-code|gemini> [--verbose]
apx install <codex|claude|claude-code|gemini> --full [--verbose]
Default manual install:
skills/ -> <agent-root>/skills/<agent-root>/plugins/<agent-root>/extensions/Give your agent access to this repo and ask it to run:
apx list
apx profiles list
apx setup <codex|claude-code|gemini> --mode recommended --yes
Agent will inspect available skills/plugins, propose a plan, and apply it. This is the legacy compatibility path for agent-curated setup; manual setup should use apx install <agent>.
Agent setup docs:
python scripts/validate-skills.py
python scripts/validate-catalog.py
python scripts/validate-mirrors.py
python scripts/check-requirements.py
The shipped catalog changes often, so it is not enumerated here. Browse it from the CLI — these surfaces are always current:
apx list # compact human browse of everything shipped
apx list --type skill # filter by category (skill, command, plugin, hook, mcp-config, ...)
apx discover "<your task>" # task-based: what should I use for this?
apx plugins list # available plugins
apx mcp list # available MCP configs
Human-facing categories: skills, commands, plugins, MCP configs, hooks, and
AGENTS.md templates (scripts, examples, and workflows are internal). See the taxonomy and
field definitions in docs/catalog-schema.md. Catalog assets also carry
a browsing tier: core, common, specialized, or experimental.
Compatibility claims in this repo are intentionally narrow:
| Asset class | Shipped today | Compatibility claim |
|---|---|---|
Root skills/ | yes | Generic text-based skills; some also mention known agent surfaces |
mcp/ | yes | MCP configs for local and remote servers (GitHub, Supabase, Vercel, Cloudflare, Exa, Atlassian, Browserbase, E2B, and more); github-local has a full check/smoke/install flow |
agents-md/ | yes | Plain text templates |
commands/ | yes | Review-first markdown command prompts; Claude Code and Codex targets where provided |
hooks/ | yes | Documentation recipes only; not installed automatically |
workflows/ | yes | Plain text scenario guides |
plugins/ | yes | Plugins with native install, marketplace metadata, and apx plugins inspection |
scripts/ | yes | Generic Python scripts |
examples/ | yes | Plain text setup examples only |
More detail: docs/compatibility.md
Most shipped skills are pure text and need no extra installation.
Current optional external tools used by shipped skills:
| Skill | Tool | Required | Install |
|---|---|---|---|
ask-codex | Codex CLI (codex) | yes for local advisor workflow | install/configure Codex CLI |
ask-claude | Claude Code CLI (claude) | yes for local advisor workflow | install/configure Claude Code CLI |
ask-gemini | Gemini CLI (gemini) | yes for local advisor workflow | install/configure Gemini CLI |
markitdown-file-intake | Microsoft MarkItDown (markitdown) | yes for conversion workflow | python -m pip install markitdown |
defuddle | Defuddle CLI (defuddle) | yes for Defuddle workflow | npm install -g defuddle |
FAQ
agent-powerups is a Claude Code plugin with 113 hand-picked skills for development work, indexed on Flowy. Install it with the command on its page. It includes prompt-evaluation-runner, red-team-eval-authoring, skill-evaluation-workbench. Its skills do not fire on their own yet. Request auto-invocation to have Flowy route them as you prompt. Free and open source.
Is this plugin yours?
Claim it with GitHubSubmit a pluginPromote it