ClaudeMap

Agentic Awesome Skills

A 2,000+-skill catalog that grew a control plane: AAS Core lets your agent search the local catalog read-only and validates the exact stack it composes.

Skills & Commandsawesome-listskillsdirectorymcpcatalog
Visit projectsickn33/agentic-awesome-skills

What began as a community awesome-list is now a two-layer system. The catalog layer holds 2,121 reusable SKILL.md playbooks plus plugins, bundles, workflows and direct installers. On top sits AAS Core (V17.1.0, agent-first preview): a local, read-only MCP server whose tools — search_skills, get_skill, list_skill_files, read_skill_file, compose_stack, inspect_stack, diff_stack — let Codex or Claude (not AAS) inspect your project, search the complete catalog and compose an exact selection. Core validates that selection in memory without ranking or recommending; a client or the aas CLI persists it as aas-stack.json with a separate selection-evidence sidecar, and `aas stack plan` produces an immutable preview with no target changes. Apply and recovery remain experimental, outside the supported preview path. An independent community project that explicitly states it is not affiliated with Google, despite its wide agent compatibility.

What it does

  • 2,121 skills in the local catalog (registry-sync header, refreshed 2026-09-12), searchable and readable file-by-file through the read-only MCP
  • Agent-owned composition: the agent searches, compares and selects; Core validates IDs and structure without ranking, recommending, excluding or hiding anything
  • aas-stack.json manifest (technical maximum of 128 skills) plus a selection-evidence sidecar; `aas stack validate` checks it, `aas stack plan` previews immutably with no target changes
  • Workbench review surface imports stack and plan JSON in browser memory only — no filesystem access, no installs; direct installs, plugins, bundles and the full catalog all remain available

When to use it

Letting your agent compare candidates across a full local catalog and commit to an exact, reviewable skill set for a project — then keeping that stack reproducible from a manifest.

Getting started

# per-agent direct install flags (--codex, --cursor, --gemini, --claude, --antigravity/--agy, ...):
npx agentic-awesome-skills --claude

# pick specific skills, dry-run first:
npx agentic-awesome-skills --skills brainstorming,systematic-debugging --dry-run

# audit an existing selection:
npx agentic-awesome-skills audit --skills brainstorming,backend-dev-guidelines

Things to know

MIT. AAS Core is agent-first-preview: catalog search and stack validation are supported, but apply and recovery are experimental; Core validates IDs and structure, not semantic fit. The npm installer requires Git 2.25+ and verifies the cloned commit against the immutable gitHead recorded for that package version.

Repository

sickn33/agentic-awesome-skills

Charted

Related on the map

  • SureForge

    Skill

    An instruction-only skill that makes agents research before they ask, plan before they build, and verify before they call anything done.

    • workflow
    • verification
    • quality
    • skill
    View

  • guizang-product-video-skill

    Skill

    Guizang's agent skill turns your repo's recent updates into a finished promo video — real components, code-rendered animation, an original code-written soundtrack — no video editor involved.

    • agent-skill
    • video
    • promo
    • marketing
    View

  • geo-sleuth

    Skill

    An agent skill that geolocates a photo with nothing readable in it — OSM geometry, elevation skylines and satellite tiles turn “where was this taken?” into a measured answer with an error radius.

    • agent-skill
    • geolocation
    • osint
    • openstreetmap
    View