Skip to content
Docs
GitHub

Use in AI Tools

Install the Everruns plugin in Claude Code, Codex, Cursor or Gemini CLI so your coding agent can build, run, debug and ship Everruns agents.

4 min readEdit on GitHub

The Everruns plugin lets a coding agent work with Everruns for you. It connects the Everruns MCP server at https://app.everruns.com/mcp (you sign in with OAuth in the browser) and adds skills that load only when a task needs them:

SkillYour agent can
everrunsUnderstand product concepts, find documentation and Cloud, and choose the right workflow skill
everruns-cliUse the CLI and session shell, discover commands, script operations, and verify changes
everruns-mcpUse an external MCP connection, run sessions, handle questions and connections, and read resources
everruns-agent-authoringDesign and revise agent behavior, harnesses, capabilities, identities, packages, and verification
everruns-run-agentHand a task to an existing agent, follow it, exchange files, and bring back the result
everruns-build-agentBuild or improve an agent from a plain-language brief, add knowledge and tools, and smoke-test it
everruns-debug-sessionDiagnose failed, stalled, or poor-quality sessions and verify fixes
everruns-evaluate-agentBuild regression evals, compare models, and score live traffic
everruns-ship-agentExpose agents through channels and triggers, with budgets and controls for stopping them
everruns-operateInvestigate organization spending, health, and change history
everruns-agents-as-filesKeep agent packages in a repository and validate, compare, import, or export them
everruns-sdkCall hosted agents from Python, TypeScript, or Rust applications
everruns-frameworkDevelop Rust applications that run agents themselves and compose advanced execution hosts

In the Everruns user menu, choose Connect your coding agent, copy the setup prompt and paste it into your coding agent. It reads the maintained plugin instructions and guides you through installation and browser sign-in. The same dialog is available after organisation setup and on an empty Agents page.

In Settings → AI clients, you can refresh approvals or revoke access. It records OAuth approval and last use, not local plugin installation.

For manual installation:

Claude Code:

/plugin install everruns --marketplace everruns/plugins

On Claude Code older than 2.1.275, run /plugin marketplace add everruns/plugins first, then /plugin install everruns@everruns.

Codex:

Terminal window
codex plugin marketplace add everruns/plugins
codex plugin add everruns@everruns

Or open /plugins in Codex and pick Everruns. Start a new session afterwards.

Everruns Codex plugin

Cursor: open Settings > Plugins, paste https://github.com/everruns/plugins, and install Everruns.

Gemini CLI:

Terminal window
gemini extensions install https://github.com/everruns/plugins

Other agents: hosts that read the Agent Plugins format can load plugins/everruns/ from the repository directly. Any MCP client can add https://app.everruns.com/mcp as a remote server named everruns.

  1. Ask your agent “Who am I on Everruns?” and complete the sign-in in your browser.
  2. Ask for real work, for example “create a Hacker News summarizer agent and run it” or “why did this session fail?” with a session link.

Every change the agent makes is recorded in the entity’s change history with the reason it gave.

Your coding-agent subscription does not configure model access inside Everruns. Running Everruns agents, including Platform Chat, requires a model provider in Models → Providers. An existing Cloud-managed provider can satisfy that need. If you skip provider setup, you can still connect your coding agent for setup and management; onboarding keeps an Add a model provider action visible.

The plugin’s skills speak the same commands as the everruns CLI. When the CLI is installed and signed in, the agent runs commands there directly, which is faster than going through MCP.

The plugin points at the Everruns demo instance. For your own deployment, keep the plugin’s skills and register your server’s MCP endpoint under the same name, for example with Claude Code:

Terminal window
claude mcp add --transport http everruns http://localhost:9300/mcp

For the CLI, use everruns --api-url http://localhost:9300/api login or set EVERRUNS_API_URL. The Docker Compose stack serves both on port 9300.

The documentation site publishes itself as plain Markdown for agents and other tools that would rather read text than HTML.

Every page in those files begins with its title, its description, and a Source: line holding the canonical URL. Cite that URL, not the text file.