aegis-orchestrator

The aegis CLI and daemon binary for the 100monkeys.ai AEGIS platform — a secure, serverless runtime for autonomous AI agents. For the complete command reference, use the canonical docs.

Install

cargo install aegis-orchestrator

Or build from source:

git clone https://github.com/100monkeys-ai/aegis-orchestrator
cd aegis-orchestrator
cargo build --release -p aegis-orchestrator

Quick Start

# 1. Start the daemon
aegis daemon start

# 2. Deploy an agent
aegis agent deploy agent.yaml

# 3. Run a task
aegis task execute my-agent --input "Summarise the README"

# 4. Use machine-readable output for automation
aegis --output json agent list
aegis --output yaml workflow describe my-workflow

# 5. Stream logs
aegis agent logs my-agent

# 6. Stop the daemon
aegis daemon stop

Output Formats

Scriptable commands support a global --output <text|table|json|yaml> flag. Use json or yaml for automation, and keep the default text/table output for operator workflows.

Examples:

aegis --output json daemon status
aegis --output json task list
aegis --output yaml config show

Streaming and interactive commands remain text-only in this pass, including: aegis init, aegis up, aegis down, aegis restart, aegis uninstall, aegis task logs, aegis agent logs, and aegis workflow logs.

aegis config generate now uses --out <path> for the destination file:

aegis config generate --out ./aegis-config.yaml

Command Surface

Implemented top-level commands: daemon, task, node, config, agent, workflow, update, init, down, up, restart, status, uninstall.

aegis node leave currently exists in the CLI surface, but it returns the baseline-protocol error and should not be documented as a working workflow.

Use the docs reference for current subcommands, flags, and behavior.

Debug Logging

# Enable verbose structured logs (includes bootstrap.py stdout/stderr)
RUST_LOG=debug aegis daemon start

# Or via config (aegis-config.yaml)
spec:
  observability:
    logging:
      level: debug   # trace | debug | info | warn | error

Configuration

The daemon is configured via aegis-config.yaml in the working directory:

apiVersion: 100monkeys.ai/v1
kind: NodeConfig
metadata:
  name: my-node
spec:
  node:
    id: my-node-001
    type: edge
  max_execution_list_limit: 1000
  llm_providers:
    - name: local
      type: ollama
      endpoint: http://localhost:11434
      enabled: true
      models:
        - alias: default
          model: phi3:mini
          capabilities: [code, reasoning]
  llm_selection:
    strategy: prefer-local
    default_provider: local
  temporal:
    address: temporal:7233
    namespace: default
    task_queue: aegis-agents
    max_connection_retries: 30

See the Node Config Reference for all fields.

Documentation

Related Crates

License

Copyright © 2026 100monkeys AI, Inc.

Licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).