This is the package manager for your Claude agent. When a user asks "how do I do X" or needs a capability you don't have built in, this searches Alibaba Cloud's skill marketplace and offers installations. It hits a Python API to search, presents results with descriptions and repo links, then runs npx skills add commands after confirmation. The bilingual trigger list and sandbox emphasis suggest it's aimed at Chinese and international users who need safe execution environments. Honest take: it's meta tooling that makes the agent extensible, which is clever, but you're trusting marketplace quality and adding dependencies at runtime. Works best when you actually need specialized tools rather than general problem solving.
npx -y skills add agentbay-ai/agentbay-skills --skill find-skills --agent claude-codeInstalls into .claude/skills of the current project.
This skill allows you to explore and install new capabilities into the agent environment. It acts as the "package manager" for your agent's skills.
Use this skill when the user:
scripts/request.py with POPRequest class
UserQuery (search term), TopK (number of results)npx skills add <repo> -s <skill_name>
-s, --skill <skills...> to specify which skill(s) to install from the reponpx skills add owner/repo -s skill_namenpx skills add https://github.com/owner/repo -s skill_namenpx skills add https://github.com/owner/repo/tree/main/skills/skill_namenpx skills add https://gitlab.com/org/repo -s skill_namenpx skills add git@github.com:owner/repo.git -s skill_namenpx skills add ./my-local-skills -s skill_name-s '*' to install all skills from the repoWhen a user asks for help with something, identify:
The search process uses the Python request module to query the skill marketplace:
Convert user's natural language intent into a concise search keyword (maintaining the original language).
from scripts.request import POPRequest
import json
# Create request
request = POPRequest(
http_method='GET',
canonical_uri='/',
x_acs_action='SearchPublicMarketSkill',
host='wuyingai.cn-shanghai.aliyuncs.com',
base_url='wuyingai.cn-shanghai.aliyuncs.com',
x_acs_version='2026-01-08',
signature_version='1.0'
)
# Search for skills
request.set_query_params({
'UserQuery': 'sandbox', # User's search query
'TopK': '10'
})
# Execute and parse
response = request.call()
results = json.loads(response)
Example Searches:
When you find relevant skills, present them to the user with:
Use the returned SourceUrl (repo URL) and SkillName from the search results to format the install command:
npx skills add <source_url> -s <skill_name>
Example response:
I found a skill that might help! The "vercel-react-best-practices" skill provides
React and Next.js performance optimization guidelines from Vercel Engineering.
To install it:
npx skills add https://github.com/vercel-labs/agent-skills -s vercel-react-best-practices
Learn more: https://pre-agentbay.console.aliyun.com/agentbay-skills/skill-detail?skillId=hw8edas7skphamvk
Only after user says "yes", "install", or similar confirmation, install the skill:
npx skills add <source_url> -s <skill_name>
When searching, consider these common categories:
| Category | Example Queries |
|---|---|
| Web Development | react, nextjs, typescript, css, tailwind |
| Testing | testing, jest, playwright, e2e |
| DevOps | deploy, docker, kubernetes, ci-cd |
| Documentation | docs, readme, changelog, api-docs |
| Code Quality | review, lint, refactor, best-practices |
| Design | ui, ux, design-system, accessibility |
| Productivity | workflow, automation, git |
| Sandbox | agentbay-aio-skills |
vercel-labs/agent-skills or ComposioHQ/awesome-claude-skillsIf no relevant skills exist:
Example:
I searched for skills related to "xyz" but didn't find any matches.
I can still help you with this task directly! Would you like me to proceed?
If this is something you do often, you could create your own skill:
npx skills init my-xyz-skill
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills