Free tools Windows power users keep installed
One-click scans. No signup required.
An Agent Skill is a reusable folder of instructions and optional resources that helps a compatible AI agent carry out a particular task or workflow. To use one, make it available in the agent’s supported environment; the agent can then discover it from its name and description, load its SKILL.md when relevant, and consult any supporting files as needed. The format is shared, but installation and runtime behavior depend on the product.
What an Agent Skill contains
A Skill packages procedural knowledge for an agent: how to perform a task, what conventions to follow, and which supporting resources may help. The folder is centered on SKILL.md, which contains metadata and instructions. It may also include references, scripts, templates, and other resources. The format documentation describes this as an open standard developed by Anthropic, intended for reuse across compatible products: Agent Skills format documentation.
![The Microsoft Copilot Bible: [3 in 1] A Beginner's Guide to Harness AI in Office 365 to Automate Workflows, Generate Content, Analyze Data, and 10x Your Productivity](https://m.media-amazon.com/images/I/41+5Qv4pd2L._SL160_.jpg)
As an Amazon Associate I earn from qualifying purchases.
OpenAI describes Skills as modular instructions for codifying processes and conventions, such as style guides or multi-step workflows. A Skill provides guidance and resources; the platform integration determines how an agent finds it and whether it can execute its instructions or bundled code.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
How an agent finds and uses a Skill
Many implementations use progressive disclosure so the full instructions do not need to be loaded for every task. The general sequence is:
#1 Best Overall
- Discovery: The agent is shown the Skill’s name and description.
- Activation: If the current task matches the description, the agent reads the instructions in
SKILL.md. - Execution: The agent follows those instructions and, when appropriate and supported, reads references or uses bundled resources such as scripts.
This means the description is functional, not just a label. It should tell the agent what the Skill does and when it applies, using language that reflects real tasks a user might request. Vague or overly broad descriptions can make it harder for the agent to select the right Skill. See Anthropic’s explanation of progressive disclosure and Skills and its Skill authoring guidance.
How to create a basic Skill
- Create a folder and a
SKILL.mdfile. Put the Skill’s name, description, and procedural instructions in the file. Follow the target platform’s current format requirements. - Write a specific description. State the task the Skill handles and the circumstances in which the agent should use it. Avoid descriptions that could match almost any request.
- Make the instructions actionable. Describe the workflow clearly, including relevant inputs, expected outputs, and important constraints. Keep the core procedure in
SKILL.md. - Add supporting files only where useful. If the workflow needs substantial reference material, repeatable actions, or reusable templates, place them in supporting files. OpenAI’s guide suggests separating these into directories such as
references/,scripts/, andassets/. - Test it in the target environment. Check whether the agent discovers the Skill for an appropriate request, follows the intended procedure, and can access any required resources. The official repository recommends testing in the environment where the Skill will be used, particularly before relying on it for critical work.
For the documented Agent Skills format, the name is lowercase and hyphenated, with a maximum of 64 characters; the description must be nonempty and no longer than 1,024 characters. These limits are format-specific guidance, so check the current rules of your target platform before depending on them. See the format’s authoring documentation.
my-skill/
SKILL.md
references/
scripts/
assets/Make the Skill available in the product you use
A folder that follows the shared format is not automatically installed everywhere. Use the instructions for the specific agent surface, and confirm its current support for the format and any files or scripts your Skill needs.
Recommended Free Tools
- OpenAI API: OpenAI documents creating Skills from a directory or ZIP upload, attaching them to hosted shell tools, or supplying them inline as a ZIP bundle. Its documentation also covers versioned bundles and validation limits. Start with the OpenAI Skills guide.
- Claude API: Anthropic documents using Skills with the code execution tool, selecting them by Skill ID, and uploading custom Skills through the Skills API. See the Anthropic Skills documentation.
- Claude Code: Anthropic documents filesystem-based custom Skills in personal
~/.claude/skills/and project.claude/skills/locations. Its public repository also describes plugin-marketplace installation for its example and document Skill sets. Consult the same Anthropic documentation for the current supported workflow. - Claude.ai: Anthropic documents prebuilt and uploaded custom Skills, with plan and code-execution requirements for custom Skills. Custom Skills do not automatically sync among Claude.ai, the API, and Claude Code; check the current product documentation for availability.
These are examples, not interchangeable installation steps. A compatible format does not guarantee that every platform supports every feature, resource type, or runtime behavior.
Rank #3
Check permissions and trust before enabling one
Review the instructions and code in a Skill before making it available to an agent. Instructions can affect planning and tool use, and scripts may perform actions in the environment where they run. OpenAI specifically warns that Skills can introduce prompt-injection-driven data-exfiltration risk; treat an unfamiliar Skill as untrusted until you understand its provenance and behavior. See the OpenAI security guidance.
Runtime permissions vary by product, so assess what a Skill expects before using it. Anthropic documents that Claude API Skills run in a sandbox without network access or runtime package installation, whereas Claude Code Skills have the network access of other programs on the user’s computer. A workflow that needs external services, packages, or local files may therefore behave differently across those surfaces. Check the Anthropic Skills documentation for the environment you intend to use.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What portability does—and does not—mean
The open format gives authors a common way to package instructions, but portability is not a promise of identical installation or execution. Before sharing or moving a Skill, verify that the destination supports its format, that the expected files can be accessed, and that its runtime permissions meet the workflow’s needs. Availability, upload methods, and supported features can change, so consult current platform documentation rather than assuming one product’s setup applies to another.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsQuick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
