Skills Overview
Skills are extensions that expand what your agents can do. They add new tools, integrate with third-party services, and define reusable workflows. Aigentic offers three ways to get skills: the ClawHub catalog, SkillForge, and manual creation.
What Are Skills?
Section titled “What Are Skills?”A skill is a packaged extension that adds specialized capabilities to an agent. Skills go beyond the built-in plugins to provide targeted functionality.
Examples of what skills can do:
- New tools — Give agents the ability to interact with databases, create charts, manipulate images, or call specific APIs.
- Service integrations — Connect agents to services like GitHub, Jira, Google Workspace, Notion, or any REST/GraphQL API.
- Workflows — Define multi-step processes that an agent can execute as a single action.
The Skills Page
Section titled “The Skills Page”Open Skills from the sidebar. The page has three tabs:
| Tab | Purpose |
|---|---|
| Browse | Search and install skills from the ClawHub catalog |
| Installed | View, enable/disable, edit, update, and remove your installed skills |
| Create | Generate skills with SkillForge (AI-powered) or build them manually with the SkillCreator wizard |
ClawHub Catalog (Browse Tab)
Section titled “ClawHub Catalog (Browse Tab)”The Browse tab is the built-in directory of available skills. It offers:
- Search — Search skills by name, tag, author, or category
- Category filter — Filter by category (productivity, development, research, communication, automation, data, security, creative, finance, utilities, integrations)
- Trust level filter — Filter by Verified, Community, or Flagged status
- Sort — Sort by Most Popular, Highest Rated, Recently Updated, or Alphabetical
- Tag quick-filter — Click popular tags to narrow results (AND logic — skills must match all selected tags)
- Featured skills — A curated section shown when no filters are active
Each skill card shows the skill’s emoji, name, author, trust level badge, description, download count, rating, category, and tags. Click a card to open a detail dialog with full information, security reports, requirements, permissions, and install/report actions.
Trust Levels
Section titled “Trust Levels”| Level | Badge | Meaning |
|---|---|---|
| Verified | Green shield | Audited and verified by the platform |
| Community | Blue users icon | Community-contributed, not yet verified |
| Unverified | Yellow shield | New or unreviewed |
| Flagged | Red shield | Has security warnings — review before installing |
Skills with critical security flags are blocked from installation.
SkillForge
Section titled “SkillForge”SkillForge is Aigentic’s AI-powered skill generator. Describe what you want in plain English and SkillForge generates a complete, functional skill package.
How to Use SkillForge
Section titled “How to Use SkillForge”- Open Skills from the sidebar and go to the Create tab.
- Describe what you want the skill to do in natural language. For example:
- “A skill that fetches the latest cryptocurrency prices from CoinGecko”
- “A skill that translates text between languages using DeepL”
- “A skill that searches Hacker News for trending stories”
- Select a model for generation (defaults to DeepSeek V3.2; only balanced and powerful models are available).
- Click Forge Skill. The generation runs through 4 phases: Extracting → Expanding → Generating → Validating, with progress bars and live file preview.
- Review the generated skill — SkillForge produces up to 7 instruction files covering foundations, components, examples, dashboards, interactive guides, and recipes.
- Click Install Skill to add it to your library, or Regenerate to try again.
SkillForge is the fastest way to create custom skills without writing code.
How Skills Work
Section titled “How Skills Work”When an agent has a skill installed:
- The skill’s tool definitions are loaded into the agent’s available tools.
- During a conversation, the model can invoke skill tools just like built-in tools.
- Skill tool invocations are routed through the gateway to the skill executor.
- Results are returned to the model for incorporation into the response.
Skill actions are logged to the audit log just like built-in tool invocations.
Skills vs Plugins
Section titled “Skills vs Plugins”| Aspect | Plugins | Skills |
|---|---|---|
| Source | Built-in or installed from npm, local path, or archive | Installed from ClawHub catalog or created with SkillForge/SkillCreator |
| Scope | Gateway (all requests) or Per-Agent (assigned individually) | Per-Agent (attached individually) |
| Examples | Memory, Web Search, Rate Limiter, CORS | Custom API integrations, workflows |
| Configuration | Toggle on/off, configurable settings with engine restart | Install, then attach to agents |
Next Steps
Section titled “Next Steps”- Installing Skills — How to find, install, and manage skills
- Creating Skills — Build your own custom skills