Recommended Free Tools
To create a Claude Skill, make a folder containing a SKILL.md file, describe the Skill’s purpose and when Claude should use it in YAML frontmatter, then install it on the Claude surface you use. Claude.ai takes a ZIP upload, Claude Code discovers Skill folders on disk, and the API has its own file-upload route. Test the result with several prompts rather than assuming it will activate for every related request.
What a Claude Skill contains
A Skill is a folder organized around SKILL.md. Start with one focused, repeatable workflow—not a collection of unrelated tasks. Anthropic’s Agent Skills overview describes Skills for Claude Code and other surfaces; the Help Center’s How to create custom skills, dated July 22, 2026, explains the file and metadata structure.
![Claude AI for Beginners Bible: [5 in 1] The Ultimate Guide to Automate Your Work, Save Hours Every Week, and Use AI for Real-World Results](https://m.media-amazon.com/images/I/415+fSJacsL._SL160_.jpg)
As an Amazon Associate I earn from qualifying purchases.
For example, a meeting-summary Skill could use this structure. The folder name is illustrative, not a required Anthropic name:
Free tools Windows power users keep installed
One-click scans. No signup required.
meeting-summary/
└── SKILL.mdPut the Skill’s name and description in YAML frontmatter at the beginning of SKILL.md. The description should state both what the Skill does and the circumstances in which Claude should use it. In the body, write the actual workflow: the steps Claude should follow, the expected output, and any constraints or examples that help it do the task consistently.
#1 Best Overall
---
name: meeting-summary
description: Summarizes meeting notes into decisions, action items, and open questions. Use when asked to turn meeting notes or a transcript into a concise meeting summary.
---
# Meeting summary
1. Identify the meeting's purpose and key decisions.
2. List action items with an owner and deadline only when the notes provide them.
3. Separate unresolved questions from confirmed decisions.In the Help Center guide dated July 22, 2026, Anthropic states that the name is limited to 64 characters and the description to 200 characters. Product requirements can change, so check the current guide if validation rejects your metadata. Keep the description concise enough to make the intended trigger clear; put detailed operating instructions in the body.
Add reference documents or executable files only when the workflow needs them. A Skill can be a short instruction file or a larger package, but extra files increase what you need to review and maintain.
Choose the installation route for your Claude surface
The right route depends on where Claude should use the Skill. These options are not interchangeable: Claude.ai uses a ZIP upload, Claude Code discovers filesystem directories, and the Skills API accepts a developer-oriented file upload.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →| Surface | Installation method | Scope or prerequisite |
|---|---|---|
| Claude.ai | ZIP the Skill folder and upload it through the Skills settings. | Anthropic’s Help Center says custom Skills are available on Free, Pro, Max, Team, and Enterprise plans with code execution enabled. Organizational controls vary by plan. |
| Claude Code | Place the Skill directory in a supported filesystem location. | ~/.claude/skills/ is the personal location; .claude/skills/ is the project location. |
| Skills API | Upload files that share a top-level directory, with SKILL.md at that directory’s root. |
A developer workflow that requires the API’s current authentication and version headers. |
Install a Skill in Claude.ai
- Make sure code execution is enabled for your account. Custom Skill availability and organization controls depend on plan and settings.
- Package the Skill folder as a ZIP file. Keep
SKILL.mdat the root of that folder rather than buried inside an extra nested directory. - In Claude, go to Customize > Skills and choose the create or upload option to upload the ZIP. Anthropic may change labels; if yours differ, follow the current custom Skills instructions.
- Confirm the Skill appears in your Skills list, then enable it using its toggle.
Claude.ai’s ZIP workflow is different from the API upload. Do not use API instructions as a substitute for the product settings flow.
Install a Skill in Claude Code
For a personal Skill, put its directory under ~/.claude/skills/. For a project Skill, put it under .claude/skills/ in the project. Each Skill directory should contain its own SKILL.md, for example:
~/.claude/skills/meeting-summary/SKILL.mdUse the personal location when the Skill is for your own Claude Code work generally; use the project location when it belongs with a particular repository. Anthropic documents these locations in its Agent Skills overview. Consult the current Claude Code Skills documentation if paths or discovery behavior differ in your version.
Upload a Skill through the API
For the API route, organize the files under one top-level directory and place SKILL.md at that directory’s root. The create endpoint accepts the files and returns a Skill object. This is a developer workflow, not the Claude.ai ZIP installation process. Before making a request, follow the current authentication and version-header requirements in Anthropic’s Create Skill API reference.
Test whether Claude uses the Skill as intended
Anthropic’s Help Center advises: “After uploading a custom skill, test it with a few different prompts to ensure it works as expected.” The recommendation appears in How to create custom skills and the Use skills in Claude guidance.
Best Value
A useful first check is to try three kinds of request:
- Clear match: Give Claude a realistic input that plainly needs the Skill’s workflow, such as meeting notes for the meeting-summary example. Check whether Claude produces the expected sections and follows the instructions.
- Variation: Try a different but still in-scope input, such as rough notes rather than a polished transcript. Look for gaps in the instructions that make the result inconsistent.
- Near miss: Ask for something related but outside the Skill’s stated purpose. Check whether Claude avoids applying the workflow when it is not appropriate.
If activation or output is off, make the description more specific about the trigger or revise the body instructions, then repeat the checks. Uploading a Skill does not guarantee it will activate on every request that seems related.
Review a Skill before enabling one from someone else
A valid Skill format is not a safety guarantee. Anthropic warns that Skills from less-trusted sources can expose users to prompt injection or data exfiltration. Before enabling one, inspect its instructions and included files. Pay particular attention to code, dependencies, bundled resources, and requests to connect to untrusted network sources. Anthropic’s Use skills in Claude guidance outlines these risks.
Quick Recap
- Read
SKILL.mdand note instructions that ask Claude to reveal, transmit, or act on sensitive information. - Review any scripts, dependencies, and bundled reference files rather than assuming they are harmless because they are packaged with a Skill.
- Check whether the Skill asks to access external services or untrusted network locations, and decide whether those connections are necessary and trustworthy.
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.
