cargo-ai Skill for Claude Code: Build and Deploy Agents in Cargo
Lets Claude create and configure AI agents inside Cargo: pick the model, write the prompt, attach knowledge, connect MCP servers and deploy a release.
- Installs
- 10.6K
- on skills.sh
- Stars
- 21
- on GitHub
- Files
- 6
- besides SKILL.md
- Updated
- Oct 2026
- last repo commit
Claude uses cargo-ai ai commands to create an agent, edit its draft release (model, temperature, prompt, actions, knowledge, MCP clients), deploy it, and manage its memories. A release is a versioned snapshot, and agents run against the deployed one.
It also covers MCP in two directions: publishing a workspace MCP server your tools can call, and connecting an agent to someone else's server. Knowledge files are uploaded with cargo-content, and running an agent over records is cargo-orchestration. It needs the Cargo CLI and a Cargo sign-in.
Good for
- GTM engineers building research or scoring agents in Cargo
- Teams who want an agent grounded in their own docs (RAG)
- Developers who want to publish a Cargo MCP server to Claude, Cursor or ChatGPT
Install cargo-ai
Claude Code
- 1
Copy this prompt.
Install the cargo-ai skill in Claude Code. Run this command: /plugin marketplace add getcargohq/cargo-skills /plugin install cargo@cargo-skills Then check that it loaded and tell me how to use it.
- 2
Paste it into Claude Code. It shows the change and asks before running it.
Every Cargo skill drives the cargo-ai CLI, so install it and sign in first. A new account starts with 100 free credits and no card (README). Pick one channel, not two: the plugin and skills add both register the skills, so using both duplicates them. The whole bundle is the route the README documents; this skill's own files point to files in sibling skills such as ../cargo/references/prerequisites.md. The plugin needs Claude Code 2.1.154 or newer and adds hooks and two subagents; see the safety section.
From the project README. Source.
What's in the skill
When it loads
Build and configure AI agents inside Cargo — create an agent, choose its model and temperature, write its prompt, attach knowledge for retrieval (RAG), connect MCP tool servers, manage memories, and deploy releases. Triggers: "create an agent", "make an agent that", "give the agent our docs", "attach this knowledge base", "attach this library to the agent", "add resources to the agent release", "connect an MCP server", "expose our tools as an MCP server", "use Cargo from Claude Desktop or ChatGPT", "change the agent model", "what does the agent remember", "deploy the agent", "the agent is answering wrong". Skip when: uploading the knowledge files themselves — use cargo-content; sending the agent a message or running it over records — use cargo-orchestration.
Sections in SKILL.md
- Bootstrap
- Discover resources first
- Quick reference
- Agents
- Releases
- Coding agents (--harness-slug, --repository)
- Structured output & heartbeat — not yet exposed as CLI flags
- Templates
- Model and temperature guidance
- Knowledge for RAG (files & libraries)
- Attaching knowledge to an agent
- Letting an agent write files
- MCP — two directions, don't mix them up
- Publishing a workspace MCP server
- Serving it to a coding agent — cargo-ai mcp
- Consuming an external MCP server
Key facts
- Version
- 2.8.1
- Installs on skills.sh
- 10,554 (Nov 2, 2026)
- GitHub stars
- 21 (Nov 2, 2026)
- Last commit
- Oct 31, 2026
- Cargo CLI license
- UNLICENSED on npm; the skills bundle itself is MIT
What it includes
Creates, edits or removes an agent and moves it into a folder.
Delete ·
agent create / update / removeSets the model, temperature, prompt, actions, knowledge and MCP clients on the draft.
Runs code ·
release update-draftMakes the draft release live; harness options run a coding agent against a repo.
Runs code ·
release deploy-draftLists agent templates with full detail; there is no template get.
Runs code ·
template listAttaches files and libraries from cargo-content through the draft release resources.
Runs code ·
Knowledge for RAGCreates a workspace MCP server from tools, agents, model views and files.
Delete ·
mcp-server (publish)Serves the platform or a curated MCP server to a stdio client using your sign-in.
Runs code ·
cargo-ai mcpConnects an agent to an external MCP server with optional disabled tools.
Runs code ·
mcp-client connectReads, edits or deletes what an agent remembers, by agent or user scope.
Delete ·
memory list / update / removeLists and reads chats, votes, documents and suggested follow-ups.
Runs code ·
Chats and prompt helpers
From the official docs
Permissions
Medium risk
It can deploy live agents and expose tools to other apps through MCP. Keep published resources read-only and check the workspace before deploying.
What it can do
- Read your data
- Create and change things
- Delete things
- Read-only mode
Watch out for
Deploying makes an agent live
release deploy-draft puts the draft configuration into use. Check the model, prompt and attached actions first. github.com
A published MCP server exposes your tools and data
ai mcp-server create takes actions, agents and model views that any MCP client can call. SKILL.md says to keep model resources read-only unless the client should write. github.com
Coding-agent releases can push to a repository
With a harness and repository, SKILL.md says the agent acts as the Cargo GitHub app on a cargo/ branch and opens pull requests. Install the app only on repos you intend. github.com
Claude runs the cargo-ai CLI as you
The skill drives the cargo-ai CLI. Sign-in is saved to ~/.config/cargo-ai/credentials.json, and the docs say to confirm the workspace with cargo-ai whoami before any write. github.com
The plugin route adds hooks that run shell commands
The plugin wires a PreToolUse hook on Bash that auto-approves safe cargo-ai calls, and a SessionStart hook that runs npm install -g @cargo-ai/cli at the pinned version. Login, token creation, report creation, project deploy and any remove or delete still prompt. github.com
The plugin sends a session summary to your Cargo workspace
The SessionEnd hook passes the last 60,000 bytes of the transcript to claude -p to write a title and summary, then stores them in your Cargo workspace session record. Use the skills CLI route if you do not want that. github.com
Remove commands delete agents, servers and memories
agent remove, mcp-server remove and memory remove are deletes. The plugin approval hook still prompts for them. github.com
The plugin also adds two subagents
cargo-execution-planner and cargo-list-builder run on the haiku model with Read, Grep, Glob and Bash tools. The planner is described as read-only and the list builder as executing only pre-approved slices. github.com
Security advisories
No published security advisories as of Nov 2, 2026.
Technical details
| Talks to | api.getcargo.io (Cargo API). registry.npmjs.org (CLI install). app.getcargo.io (token creation) |
|---|---|
| Sign-in and keys | Sign-in by emailed code, browser (OAuth) or an API token. Browser and email sign-in save a session to ~/.config/cargo-ai/credentials.json; API tokens are workspace-scoped and the README says to store them in a secrets manager. |
| Scopes |
|
Questions
What does cargo-ai do?
It lets Claude build AI agents inside Cargo. It creates the agent, sets its model and prompt, attaches knowledge for retrieval, connects tool servers and deploys a release. The skill name is the Cargo ai domain, not a general AI helper.
How do I give an agent my docs?
Upload the files or build a library with cargo-content, then attach them in the draft release resources and deploy. SKILL.md says a file does nothing until it is attached to an agent release and deployed.
Can I use Cargo from Claude Desktop, Cursor or ChatGPT?
SKILL.md describes publishing a workspace MCP server any MCP client can call, and a cargo-ai mcp bridge for stdio clients that uses the credentials already on your machine. It also says Cargo serves a first-party MCP at mcp.getcargo.io.
Which model does it recommend?
SKILL.md pairs use cases with models and temperatures, for example low temperature for classification and scoring. It notes prices span three orders of magnitude, so read the price before picking a model for a high-volume agent.
Can an agent edit my code?
A release can be set up as a coding agent with a harness and a GitHub repository. SKILL.md says it works on a cargo/ branch, never pushes to the default branch and opens a pull request, and the Cargo GitHub app must be installed on the repo.
How do I install it?
Use the install box above. Install the whole Cargo bundle, then the Cargo CLI and sign in. The README says to pick one channel, plugin or skills add, not both.
Is it official?
Yes. It sits in Cargo's own repo (getcargohq/cargo-skills), the LICENSE names Cargo as the copyright holder and the plugin manifest lists Cargo as the author.
Does it work in the Claude app, Codex or Cursor?
The README names Claude Code, Codex, Cursor, Windsurf, GitHub Copilot and Hermes Agent. For Claude Desktop it documents a Claude Cowork plugin route with network egress turned on. The skill drives a local CLI, so there is no plain Claude app skill upload here.
Is it safe to use?
It changes live agents and can expose tools and data to other apps. Deploying a draft release makes it live, publishing an MCP server lets MCP clients call your tools, and remove commands delete agents and memories. Our advice: keep published MCP resources read-only and check the workspace name before deploying.
Information
updated Nov 2, 2026
- Repository
- getcargohq/cargo-skills
- Folder
cargo-ai
Sources
Docs, issue trackers and discussions this page is built from.
Show all 7Hide
- cargo-skills README · github.com
- cargo prerequisites.md · github.com
- cargo-ai SKILL.md · github.com
- cargo-skills LICENSE · github.com
- cargo-skills plugin.json · github.com
- session-end.sh · github.com
- session-start.sh · github.com