Skills

A skill is a reusable set of working methods in a workspace. Its main file is SKILL.md, optionally accompanied by scripts, templates, and reference material.

Add a skill to an agent and the agent uses its guidance when executing related tasks. The same method no longer has to be copied into each agent's instructions, and updates are maintained in one place.

Skills and agent instructions

Creating or importing a skill

On the workspace Skills page, choose "New skill". Three options are available:

Copying from a runtime takes a snapshot of the files at that moment. Later changes to the local files do not update the workspace version automatically; re-import or edit in Multica.

With the CLI, you can also import from a .skill or .zip file:

multica skill import --file ./review-helper.skill

If a skill with the same name already exists in the workspace, the import can stop, overwrite the existing content, save under a new name, or skip. Overwriting through import is limited to the skill's creator; a workspace admin who can edit the skill in the app still cannot overwrite it via import. To pull a newer version of a skill that is already in the workspace, use Updating an imported skill instead.

Updating an imported skill

A skill imported from GitHub, ClawHub, or Skills.sh keeps a reference to its source. To pull the latest version, use Update on the skill page (or "Update from source" in the list's row menu), or the CLI:

multica skill refresh <skill-id>

The skill is re-downloaded from its source and replaced in place:

• Preserved: the skill's identity — agent assignments, labels, creator, and creation date. Agents keep the skill without re-adding it.
• Replaced: name (if the source renamed it), description, SKILL.md, and all supporting files. Local edits to the skill are overwritten.

The skill's creator and workspace owners / admins can update from source. Skills created manually, imported from an archive file, or copied from a runtime have no hosted source and cannot be updated this way — re-import them instead.

Skill contents

Every skill contains a main file, SKILL.md. Spell out:

• when the skill applies;
• what to check before starting;
• the steps to execute;
• what form the result should take;
• when to stop and check with a member.

You can also add supporting files, for example:

SKILL.md
references/api-conventions.md
templates/review-report.md
scripts/check.sh

SKILL.md is a reserved path; a supporting file cannot use that name. Supporting files are delivered to the agent together with the main file.

Attaching to an agent

After creating or importing a skill, it still needs to be added to specific agents. Pick agents on the skill page, or manage bindings from the agent's Skills tab.

• One agent can use multiple skills;
• one skill can serve multiple agents;
• when a skill isn't needed for now, disable the binding instead of deleting the skill;
• edits take effect from subsequent runs and don't change tasks already running.

Only members who can modify the agent can add, remove, or disable its skills.

Workspace skills and repository skills

Workspace skills are stored in Multica — a fit for skills the team maintains together and adds to different agents.

Some AI coding tools also read project-level skills directly from the code repository, such as .claude/skills/ or .agents/skills/. Those files stay managed by the repository; Multica does not automatically register them as workspace skills. Support depends on the tool in use — see AI coding tools comparison.

Runtime-local skills already installed on a connected computer (Claude Code and Codex) show up in the agent's Skills tab and can be disabled per agent. Disabling likewise requires permission to modify the agent.

Edit permissions

Any workspace member can create or import a skill. The creator and workspace owners and admins can modify or delete it; other members can view and use it, but cannot overwrite its content.

Deletion is permanent and removes the skill from all agents at the same time.

Next steps

• Squads — let a leader pick the right member for the work.
• Daemon and runtimes — how skills reach the actual execution environment.
• AI coding tools comparison — how different tools support skills.