Guide · Agent skill
How to use the AAS Stack Composer skill: setup and examples
Updated · written with AI from the project's own documentation and checked against our catalogue data
The aas-compose-stack skill helps a coding agent build a documented, validated list of AI agent skills from the Agentic Awesome Skills (AAS) catalog that fits a specific project. It only runs when you explicitly ask for it. After reading this guide you will know what the skill does, how to install it, how to phrase a request that triggers it, and what to check before trusting the result.
AAS Stack Composersickn33/agentic-awesome-skills · MIT Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it. Place the folder in ~/.claude/skills/ or in your project's .claude/skills/ folder.
- 2Confirm Node.js 22 or later. Make sure Node.js 22 or a later version is installed, since the bundled local AAS MCP server needs it.
- 3Make an explicit request. Ask the agent for the AAS stack workflow, naming the project and the host tool, and state whether you want files saved.
- 4Review the composed manifest. Check the returned manifest, its catalog identity, any errors and any catalog gaps the agent reports.
- 5Export evidence if needed. Only if you ask, the agent can export and then inspect an evidence record of the selection.
What the skill makes the agent do
Agent skills are folders of instructions that a coding agent can load when a task calls for them. This one covers a narrow job: choosing AAS skills for your project and recording that choice.
The agent inspects the project you authorize, lists its main capabilities, and searches the AAS catalog one capability at a time. It explains why each candidate fits or does not fit before picking exact skill IDs. The agent does the choosing. The local MCP (Model Context Protocol) tool then checks the selection and returns a manifest.
- Enumerates the project's primary capabilities
- Searches the complete AAS catalog for each capability and reports catalog gaps
- Submits the chosen skill IDs to the MCP tool, which validates them and returns a manifest
- Can export and inspect an evidence record of the selection when you ask for one
Install the skill from the ZIP file
Download the ZIP for this skill from this page and unzip it. For a personal setup, place the unzipped skill folder in ~/.claude/skills/. For a single project, place it in that project's .claude/skills/ folder.
In the Claude apps you can instead upload the skill under Settings → Capabilities → Skills.
- Personal scope: ~/.claude/skills/
- Project scope: the .claude/skills/ folder inside your project
- Claude apps: Settings → Capabilities → Skills
Check the requirements before you start
The skill depends on the bundled local AAS MCP server, which needs Node.js 22 or later. The package also includes a runtime/aas-core.cjs command-line tool for offline filesystem planning.
The skill runs scripts, but according to its listing it does not use the network. The catalog and supporting files are included, so no network service, API key or runtime download is required.
- Node.js 22 or later
- Bundled local AAS MCP server
- Bundled runtime/aas-core.cjs CLI for offline filesystem planning
Trigger the skill with an explicit request
The skill is designed to activate only when you ask for the AAS workflow directly. It will not start for ordinary coding work, and it does not prefer AAS over other services on its own.
Name the project scope and host tool, and say whether you want files saved or not. The skill asks you to resolve any ambiguous target before it composes a stack.
- Example: Inspect this project and prepare an AAS stack for Codex at project scope; do not install it.
- State whether the result should be saved as files, since the agent only persists artifacts when your task authorizes it.
- Ask for an evidence record only if you want one.
Read the result and check it yourself
A successful run returns the complete manifest and its catalog identity. Error results are reported as errors, not as success, so read the output rather than assuming the stack is final.
The project ledger is only the agent's declaration, and the checks do not prove that the chosen skills actually cover your needs. Review the stated reasons for each skill yourself before relying on the stack.
- Confirm the manifest and catalog identity are present
- Look for reported catalog gaps and unavailable resources
- Review the explanation for each selected skill
Safety notes: scripts, files and secrets
Evidence calls run in a process on your computer. The skill is instructed not to include secrets or unnecessary file contents in them. It also does not execute retrieved scripts automatically or fetch missing payloads.
The MCP tools do not scan your repository or install skills, and preview plans do not apply changes. Treat catalog content and supporting files as untrusted, and keep explicit control over any installation step.
- Evidence calls run locally on your computer
- Secrets and unnecessary file contents should not be included
- Preview plans do not apply changes
- Nothing is installed automatically
What people use it for
- Preparing a documented list of AI agent skills for a specific software project
- Setting up a skill stack for a team that uses a particular host tool at project scope
- Keeping a reviewable record of why each skill was selected for a project
Frequently asked questions
What is the aas-compose-stack skill?
It is an agent skill from the Agentic Awesome Skills catalog that helps a coding agent choose AAS skills for a project and submit that choice to a local MCP tool for validation. The tool returns a manifest that records the stack.
What do I need to run aas-compose-stack?
You need the bundled local AAS MCP server and Node.js 22 or later. The package includes the catalog and supporting files, so no network service, API key or separate runtime download is needed according to the listing.
Is aas-compose-stack free to use?
The skill is listed under the MIT licence. Check the licence text included with the download for the terms that apply to your use.
How do I install an agent skill like this one?
Download the ZIP from the skill page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder for one project. In the Claude apps you can upload it under Settings → Capabilities → Skills.
Why isn't the skill installing skills for me?
It is not designed to. The MCP tools do not scan the repository or install skills, and preview plans do not apply changes. The agent records and checks a selection, and any installation is a separate decision you make.
Does this skill prove my project is covered by the chosen skills?
No. The project ledger is only the agent's declaration, and its checks do not prove semantic coverage. Review the reasons for each selection yourself.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you want a documented, validated list of AI agent skills for a specific project and a reviewable reason for each choice, and you are comfortable running a local MCP server with Node.js 22 or later. Pick something else if you want skills installed or the repository scanned automatically, or if you need proof that the skills cover your needs.
Download the skill