Skip to main content
The Endor Labs Agent Kit installs into Claude Code as a plugin, adding the Endor Labs setup skill and the full set of workflow agents.
1

Install the plugin

Install the endor-labs-agent-kit@endorlabs plugin from the public marketplace. In Claude Code, run.
/plugin marketplace add endorlabs/ai-plugins
/plugin install endor-labs-agent-kit@endorlabs
/reload-plugins
/agents
Start a new Claude Code session or run /reload-plugins after installing or reinstalling the plugin. If Claude Code still shows stale content for the same version, uninstall and reinstall the plugin. Run /reload-plugins and start a new session so host caches reload the agents and setup skill.
Upgrading from Endor Labs SkillsIf you previously installed the Endor Labs plugin as ai-plugins@endorlabs, it keeps working and existing installs are unaffected. New users should install the endor-labs-agent-kit@endorlabs plugin instead. Do not enable both plugin IDs in the same Claude Code profile, because they expose the same agents and setup skill.
2

Set up this machine

Ask your AI coding assistant to run the setup skill.
Use the endor-agent-kit-setup skill to check Endor Labs Agent Kit readiness.
The setup skill guides package-manager-first endorctl installation, verifies Endor Labs authentication and namespace readiness, and reports missing gh or toolchain prerequisites. It does not run scans, run endorctl host-check, edit shell profiles, auto-install gh, or install language runtimes and package managers.
3

Run your first workflow

Invoke a workflow agent with @agent-<name>. Try these example prompts.
@agent-ai-sast-triage triage AI SAST findings for this repository

@agent-sca-remediation check this repository for P0 SCA findings I can start remediating

@agent-probe-droid probe GitHub org <org> for Endor Labs monitored-branch onboarding gaps and setup prescriptions

@agent-endor-troubleshooter diagnose this Endor Labs scan failure from redacted error text and read-only tenant evidence
See the Agent Kit overview for the full agent catalog, and the safety model for how mutating agents handle approval gates.