Renders approval cards, diffs, progress bars and metrics on a dashboard instead of terminal text.
awslabs/cli-agent-orchestrator
Helps you set up and run several AI coding agents together, coordinating their work in isolated sessions.
- 16 skills
- 1.6k stars
The 5 most installed
Checks an Agent Plugin package's manifest and skills against CAO's schema limits and reports violations.
Scaffolds a new Agent Plugin package with its manifest, skills folder and optional MCP config.
Picks the best installed agent profile for a task before work is delegated.
Walks through the local build, test, lint and pre-PR checks for contributing to CAO.
The other 11 skills
Records task outcomes and distills lessons so repeated workflows improve over time.
Enables and troubleshoots CAO's fleet dashboard views inside MCP App hosts like Claude Desktop.
Saves and recalls durable facts like preferences, conventions and decisions across sessions.
Creates CAO plugins that react to lifecycle events and connect tools like Slack or Discord.
Adds support for a new CLI-based AI agent as a CAO provider.
Verifies whether a CAO session is really alive and what it actually output before reporting.
Finds agent profiles and messages, monitors or unblocks the terminals of a running CAO session.
Coordinates worker agents: assigns or hands off tasks, fans out work and collects results.
Follows the worker-side rules for finishing tasks and reporting results back to the supervisor.
Writes and runs repeatable multi-step CAO workflow scripts that fan out over many items.
Loads the official MCP Apps builder skills for creating, migrating or adding interactive views to MCP servers.