PlugWiki

bailian-managed-agent Skill: Bailian Managed Agents as Code

Lets your agent create and manage Bailian (Alibaba Cloud Model Studio) managed agents from an agents.yaml file with the bl CLI.

Installs
9.2K
on skills.sh
Stars
367
on GitHub
Files
2
besides SKILL.md
Updated
Nov 2026
last repo commit

bailian-managed-agent makes your agent treat a Bailian managed agent like infrastructure as code. It scaffolds agents.yaml, validates it, shows a plan of changes, and applies or destroys it only after you confirm. It also covers directory projects, sessions, files, memory stores, deployments and webhooks.

This is Alibaba Cloud Bailian, not Anthropic's product. It needs a Bailian API key, the bl tool and the shared bailian-protocol skill. Apply and destroy change real cloud resources.

Good for

  • Developers building cloud-hosted agents on Alibaba Cloud Bailian
  • Teams who want agent setups reviewed as a plan before they are applied
  • Ops people who run, pause and inspect deployments, sessions and webhooks

Install bailian-managed-agent

Claude Code

  1. 1

    Copy this prompt.

    Install the bailian-managed-agent skill in Claude Code. Run this command:

npm install -g bailian-cli
bl skill init

Then check that it loaded and tell me how to use it.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js 18.17+.

bailian-managed-agent needs the shared bailian-protocol skill and the bl command. bl skill init installs every bailian-* skill together with bailian-protocol, which is the supported route. Windows PowerShell: irm https://bailian.aliyun.com/cli/install.ps1 | iex. The skills CLI fallback targets every agent it finds, because of --all, and INSTALL.md pins version 1.5.18 for Node 18 support. The README does not say which agents bl skill init writes to, so check your skills folder afterwards.

From the project README. Source.

What's in the skill

When it loads

阿里云百炼托管 Agent 声明式基础设施与 API 命令入口：用户要创建agent、初始化 agents.yaml、校验或预览配置变更、 创建/更新/销毁托管 Agent 或 Deployment、在 Workbench 编辑和调试目录项目、管理本地快照版本，或查询 Agent/Environment/Skill/Vault/Deployment/Memory Store、管理 Session/Event/File/Webhook/记忆正文及版本、运行/暂停 Deployment 时使用 `bl managed-agent`。Agent/Environment/Skill/Vault/Deployment/Memory Store 仍以 agents.yaml 为唯一事实源做 IaC；Webhook 使用无需项目文件的工作空间 API；公开 API 能力按资源透出 list/get/search/versions/download、数据面和运行时动作命令。apply / destroy 与破坏性 API 命令受统一高风险确认闸门保护； 务必先展示预览再让用户确认，禁止自动添加 `--yes`。 反触发：调用已上线的百炼应用/智能体走 bailian-app-call 或 `bl app`；宿主 agent 自身的记忆、技能、 子代理不走本 skill；生图生视频走 bailian-gen。 官方安装：`bl skill init`（与共享协议 bailian-protocol 同装）。

Sections in SKILL.md

  1. Safety guardrail (the most important rule)
  2. IaC lifecycle
  3. Directory projects, Workbench, and local versions
  4. Scoped single-resource create
  5. Credential secret input
  6. Deployment as IaC
  7. Session interaction (chat with a deployed managed agent)
  8. API-oriented resource commands
  9. Memory Stores and Session mounts
  10. Workspace Webhooks
  11. Local state management
  12. Common hand-offs
  13. references

Key facts

Author
Aliyun Model Studio CLI repo (modelstudioai/cli); its README calls the CLI "the official command-line interface for Aliyun Model Studio (DashScope) AI Platform"
License
Apache-2.0 (README badge and LICENSE file)
Installs on skills.sh
9,235 (Nov 2, 2026)
GitHub stars
367 for the whole modelstudioai/cli repo (Nov 2, 2026)
Last repo commit
Nov 1, 2026
Files in the folder
3 files including SKILL.md: reference/index.md, reference/managed-agent.md
Needs
The bl command (npm package bailian-cli, Node.js 18.17+), the bailian-protocol skill and a Bailian API key or console login
Version in frontmatter
2.7.2

What it includes

  • init, validate, plan, apply and destroy work from agents.yaml

    validate is offline. plan shows the diff. apply and destroy need --yes after your confirmation.

    Try: “Create a Managed Agent that can generate short-film storyboards and videos”

    Delete · IaC lifecycle

  • project init, validate, build, publish, workbench and version commands

    Build is local. Publish needs --yes and uploads the project source, including skill scripts and assets. Version restore changes your source files.

    Write · Directory projects

  • Creates one agent, environment, skill, vault, credential, deployment or memory store

    Seven create commands. Without --yes they only preview; with --yes they write YAML and create the remote resource.

    Write · Scoped single-resource create

  • session run, send, create, get, list, events and delete

    Chat with a deployed agent, stream events and export diagnostics. archive and delete need confirmation.

    Write · Sessions

  • list, get, search and versions for agents, environments, skills, vaults and deployments

    Also deployment run, pause and unpause, and file upload, download and delete.

    Network · Resource queries

  • Creates stores and reads or edits memory contents and versions

    Version redact permanently removes a version's content and needs a dry run first.

    Delete · Memory stores

  • list, create, update, delete, enable, disable, test and reset-secret

    All writes are high-risk. test sends events to your receiving endpoint. create and reset-secret responses can hold a signing secret.

    Write · Webhooks

  • state list, show, import and rm

    state rm only edits the local state file; destroy deletes the remote resource.

    Write · Local state

From the official docs

Permissions

High risk

The skill can create, change and destroy cloud agents on your Bailian account and writes secrets to a local .env file. Read each plan, and approve apply or destroy yourself.

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • Apply and destroy change real cloud resources

    SKILL.md calls this the most important rule: run plan, show the diff, and add --yes only after you confirm. Read the plan and refuse if it deletes something you want to keep. github.com

  • Secrets land in a plaintext .env file

    Build moves Vault secret values and access tokens into the project's root .env. It is not encrypted and not git-ignored for you. Add it to .gitignore and back it up safely. github.com

  • Webhook tests send data to your endpoint

    webhook test delivers events to the receiving URL, and create or reset-secret responses can include a signing secret. Keep that secret out of logs and chats. github.com

  • Some deletions cannot be undone

    memory-store versions redact permanently removes content. state rm only untracks locally, while destroy deletes remotely. Check which command the agent proposes. github.com

  • It depends on the protocol skill and the bl CLI

    It reads bailian-protocol first, which installs software, runs bl skill update and prefixes commands with usage attribution. See the bailian-protocol page for those findings. github.com

  • Some features are China-site only

    The README says app orchestration, native MCP setup, training and deployment, account operations and plan onboarding are currently available only to China site (aliyun.com) account holders. Check your account type first. github.com

Technical details
Talks toDashScope and Bailian API endpoints via the bl CLI. Your webhook receiver URL when you run webhook test
Sign-in and keysBailian API key or console login through bl. Vault secrets in directory projects are moved into a project .env file.
Scopes

  • Runs bl managed-agent with your Bailian API key

  • apply, destroy and single-resource create change persistent cloud resources

  • Writes .env, agents.yaml and a local state file in your project

Questions

What does bailian-managed-agent do?

It makes your agent create and manage Bailian managed agents from a config file. It follows plan, confirm, apply, so you see changes before they happen.

How do I install it?

Use the install box above. Install the whole Bailian set with bl skill init, because this skill reads bailian-protocol first and stops without it.

Is this Anthropic's managed agents?

No. It is for Alibaba Cloud Bailian (Aliyun Model Studio). It is unrelated to Anthropic's agent products.

Is it official?

It is in the vendor's own repo. The README of modelstudioai/cli calls it "The official command-line interface for Aliyun Model Studio (DashScope) AI Platform".

Does it cost money or need a key?

It needs a Bailian API key or console login. Managed agents and deployments are created in your Alibaba Cloud Model Studio (Bailian) account, so that account is billed under Model Studio billing. SKILL.md gives no prices and points quota and billing questions to the bailian-cli skill (bl quota, bl usage). Check Bailian pricing before you apply a plan.

Does it work outside China?

The README says app orchestration, which includes calling Managed Agents, is currently available only to China site (aliyun.com) account holders. Check your account.

Where do secrets go?

Build moves literal Vault secret_value and access_token values into a .env file at the project root and swaps in environment references. SKILL.md says .env is private plaintext, is not encrypted and is not ignored by Git for you. Add it to .gitignore.

Is it safe?

It can create, update and destroy cloud resources and send test events to a webhook URL. The skill says never to add --yes before you confirm. Read the safety box before you apply anything.

Information

updated Nov 2, 2026

Repository
modelstudioai/cli
Folder
skills/bailian-managed-agent

modelstudioai/cliskills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 5
  1. Aliyun Model Studio CLI README · github.com
  2. Aliyun Model Studio CLI INSTALL.md · github.com
  3. bailian-protocol setup.md · github.com
  4. bailian-protocol SKILL.md · github.com
  5. bailian-managed-agent SKILL.md · github.com

Updated Nov 2, 2026.