Guide · Agent skill
How to use the Write agentmemory Skill agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The write-agentmemory-skill skill is an agent skill that teaches an AI agent the house format and rules for creating or editing a skill in the agentmemory project. It is written for contributors adding a new skill, maintainers reviewing skill pull requests, and anyone restructuring an existing agentmemory skill folder. After this guide you will know what the skill does, how to install it, how to get your agent to use it, and where its limits are.
Write agentmemory Skillrohitg00/agentmemory · Apache-2.0 Download the skill Quick start
- 1Download the ZIP. Download the skill ZIP file from this page.
- 2Unzip into a skills folder. Unzip it into ~/.claude/skills/ for a personal install, or into .claude/skills/ in a project.
- 3Or upload in the Claude apps. In the Claude apps, open Settings, then Capabilities, then Skills, and upload the ZIP.
- 4Check the files. Confirm the folder contains SKILL.md and TROUBLESHOOTING.md.
- 5Ask with a matching request. Describe a task that adds, restructures or reviews an agentmemory skill, so the description's triggers match.
- 6Review the output. Check any skill edits the agent proposes against the rules before accepting them.
What is an agent skill and what does this one do?
An agent skill is a folder containing a SKILL.md file and optional supporting files. The agent reads the short description at the top of SKILL.md to decide whether to load the rest of the skill for a task.
This skill covers the rules for writing agentmemory skills. Once it is loaded, the agent knows how SKILL.md should be structured: the frontmatter, the wording of the description, the section order, the line limits, and style rules such as no emojis and no em-dashes. It also knows that facts taken from the source code, such as tool names and endpoints, should be produced by a generator command rather than typed by hand.
- Frontmatter fields: name, description, optional argument-hint, and user-invocable
- Description written as two sentences, the second starting with 'Use when'
- Required body order: Quick start, Why, Workflow, Anti-patterns, Checklist, See also, and a Reference or Troubleshooting pointer
- SKILL.md kept under 100 lines, with dense facts moved to REFERENCE.md and examples to EXAMPLES.md
- No emojis, no em-dashes, no filler
Install the skill on Claude Code
Download the ZIP file for this skill from this page. Unzip it so that the skill folder sits inside your skills directory. For a personal install, use ~/.claude/skills/. For a single project, use the .claude/skills/ folder inside that project.
After unzipping, the folder should contain SKILL.md and TROUBLESHOOTING.md. Start a new session so the agent can discover the skill.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ in your project root
Install the skill in the Claude apps
If you use the Claude apps instead of Claude Code, you can upload the skill directly. Open Settings, go to Capabilities, then Skills, and add the ZIP file you downloaded from this page.
The skill content is the same either way. Only the place where you install it differs.
- Path in the Claude apps: Settings → Capabilities → Skills
How do you trigger the skill with a natural request?
This skill is not a slash command. Its frontmatter sets user-invocable to false, so you do not type a command to run it. Instead, the agent loads it when your request matches the triggers in the description.
The description says to use the skill when adding a new skill, restructuring an existing one, or reviewing a skill contribution for consistency. Phrase your request in those terms so the trigger matches.
- 'Add a new skill to the agentmemory repo for handling session summaries.'
- 'Restructure the existing skill folder so SKILL.md fits under 100 lines.'
- 'Review this skill pull request for format consistency with the house rules.'
What do you need before using it?
The listing does not require any tools, packages or accounts to install this skill. It does not run scripts and does not use the network.
The rules are specific to the agentmemory project. They assume its directory layout, such as plugin/skills/<name>/, and its generator command, npm run skills:gen. The skill can explain those rules, but running the generator requires the agentmemory repository itself.
- Runs scripts: no
- Uses network: no
- Projects outside agentmemory will not match its directory layout or generator
npm run skills:genWhat are the safety notes and limits?
Because the skill runs no scripts and makes no network requests, it only adds written guidance to the agent's context. It does not change your files on its own. Any edits to a skill folder come from the agent following the rules, and you should review them as you would any other change.
The skill does not provide agentmemory memory features. If you only want the agent to remember things, this skill will not help with that.
- Does not provide memory functions
- Does not edit your work directly; it is a reference for contributors
What people use it for
- Adding a new skill folder under plugin/skills/ with a correctly formatted SKILL.md
- Restructuring an existing skill so its body follows the required section order
- Reviewing a skill pull request for frontmatter, description wording and line limits
- Checking that facts in a skill come from the generator rather than being typed by hand
Frequently asked questions
Is the write-agentmemory-skill skill free to use?
The skill is listed under the Apache-2.0 licence. Check the licence file in the agentmemory repository for the exact terms before reusing the material.
What do I need to install this agent skill?
The listing does not require any tools or accounts. You download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it in the Claude apps under Settings, Capabilities, Skills.
Does this skill run scripts or use the internet?
No. The listing states that it runs no scripts and uses no network. It only gives the agent written rules to follow.
Can I use this skill outside the agentmemory project?
It is not a good fit. The rules depend on agentmemory's directory layout and its generator command, npm run skills:gen, so they will not match another project's structure.
Will this skill give my agent memory features?
No. It is a reference for writing agentmemory skills and does not provide any memory functions. Use agentmemory itself for memory features.
Why won't my agent load the skill?
The agent decides to load a skill from its description. Make sure the skill folder sits in the correct skills directory, that SKILL.md is present, and that your request matches a trigger such as adding, restructuring or reviewing a skill. The bundled TROUBLESHOOTING.md covers recovery steps for separately installed skills.
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 are contributing a skill to agentmemory, reviewing one, or restructuring an existing skill folder, because it encodes the project's format and style rules. Pick something else if you only want memory features from agentmemory or need guidance for a project with a different layout.
Download the skill