AI Skills & Plugins
AI skills and plugins extend an existing assistant or coding agent with packaged instructions, reference files, scripts or tool connections that load when a task calls for them.
90 results in AI Skills & Plugins
Active Pro listings from this category appear first. Your sort applies within Pro and Free results.
Shortlists for AI Skills & Plugins
Explore collections with tools relevant to this category, grouped by deployment, source model or workload.
Open-source IT tools →
81 matching tools from this category
Free IT tools →
5 matching tools from this category
Self-hosted IT tools →
34 matching tools from this category
How to choose AI Skills & Plugins tools
AI skills and plugins extend an existing assistant or coding agent with packaged instructions, reference files, scripts or tool connections that load when a task calls for them. A skill teaches a method, such as how to audit a site or build a slide deck, while a plugin can bundle several skills with commands and hooks. This differs from MCP servers, which expose tools and data through a protocol, and from agent frameworks, which build new agents.
Who uses these tools
- Developers customising coding agents for their projects
- Specialists packaging their methods for assistant use
- Teams sharing approved workflows across many assistant users
Capabilities to compare
- Instruction files loaded on demand by task
- Bundled scripts, templates and reference material
- Commands and hooks registered with the host
- Support across several assistant hosts
- Versioned distribution through marketplaces or repositories
Operating considerations
- Skills run with the permissions of the host assistant
- Trust in the author of any bundled scripts
- Context budget consumed by large reference files
- Portability between assistant hosts and skill formats
Selection checklist
- Pick the subcategory that matches the task, then shortlist skills with clear documentation and examples.
- Read the instruction files and scripts before installing, as you would review any third-party code.
- Test each skill in isolation so you can attribute changes in assistant behaviour to the right package.
- Prefer skills you can pin and update deliberately over ones that change silently.
What belongs in this category: Products belong here when they are installed into an assistant you already use to add domain knowledge or workflows, grouped below by field such as development, design, marketing, research, security and business.
Typical use cases
- Adding a team's coding conventions to every agent session
- Giving an assistant a repeatable audit or review method
- Turning documentation into knowledge an assistant loads on demand
- Sharing approved workflows across an organisation's assistant users
Trade-offs to model
- Skills are lightweight text and easy to audit, while MCP servers offer live tool access but add a running service to maintain.
- Community skills cover many niches for free, whereas internally written skills match your processes but need ongoing maintenance.
Common mistakes
- Treating skills as harmless text and installing bundles whose scripts or hooks were never reviewed.
- Loading so many skills that the assistant's context fills up and instructions start to conflict.
- Choosing a skill format tied to one host when the team uses several different assistants.
AI Skills & Plugins on iThub in figures
iThub lists 90 tools in AI Skills & Plugins. 81 of 90 are open source, 34 of 90 can be self-hosted, 86 of 90 have a free, freemium or open-source pricing model and 13 of 90 offer an API.
- Tools listed
- 90
- Open source
- 81
- Self-hostable
- 34
- Free to start
- 86
- With an API
- 13
- Web app
- 11
- Desktop app
- 47
Recently added: Hallmark, GitHub Spec Kit, unslop, Unslopovich
Learn how assistant skills work Explore MCP servers See agent frameworks
Continue the technical evaluation
Use the resources that match your decision stage. Profile data and linked sources were refreshed through November 21, 2026.
