Developer Tools

Astromesh gives you a complete toolkit for building, running, and monitoring AI agents. Three tools, one workflow — from YAML to production.

The Workflow

Define → Run → Debug → Optimize → Deploy
  │        │       │         │         │
  YAML   CLI    Traces    Metrics   Docker/K8s
  │        │       │         │
  VS Code IntelliSense  Dashboard

astromeshctl — The CLI

Your command center. 16 commands that cover the full agent lifecycle.

Step 1: Scaffold

astromeshctl new agent customer-support

Generates a ready-to-run YAML file in config/agents/. Open it in VS Code for instant IntelliSense.

Step 2: Run

astromeshctl run customer-support "How do I reset my password?"

Executes the agent against your local runtime. Output streams to the terminal.

Step 3: Debug

astromeshctl traces customer-support --last 5

Full execution trees — guardrails, memory lookups, LLM calls, tool usage, and timings. Every step visible.

Step 4: Monitor

astromeshctl metrics customer-support
astromeshctl cost --window 24h

Token usage, latency histograms, and cost tracking. Catch expensive patterns before they hit production.

Step 5: Diagnose

astromeshctl doctor

Checks runtime, providers, memory backends, and connectivity. One command to know if everything is healthy.

Full reference: CLI Commands

Copilot — AI That Knows Your Project

An AI assistant embedded in both CLI and VS Code. It reads your agent configs, understands your runtime, and answers questions in context.

astromeshctl ask "Why is my agent using so many tokens?"

astromeshctl ask "Add a RAG pipeline to my support agent"

astromeshctl ask "Explain the last trace for sales-qualifier"

The Copilot doesn’t just answer generic questions — it inspects your actual YAML files, runtime state, and trace history to give specific answers.

In VS Code, open the Copilot Chat panel for an interactive session with the same capabilities.

VS Code Extension — Your Editor as Mission Control

Seven features that turn VS Code into a full agent development environment.

Feature	What it does
YAML IntelliSense	Auto-complete and validation for `.agent.yaml` and `.workflow.yaml`
▶ Play Button	Run agents directly from the editor title bar
Traces Panel	Expandable span trees in the sidebar, auto-refreshing
Metrics Dashboard	Real-time counters and histograms in a webview
Workflow Visualizer	DAG visualization of workflow steps
Copilot Chat	AI assistant panel with full project context
Diagnostics	`astromesh doctor` results in the output channel

Install

Search “Astromesh” in the VS Code marketplace, or:

code --install-extension monaccode.astromesh

Full reference: VS Code Extension

Multi-Agent Workflows

When one agent isn’t enough, compose them:

tools:
  - name: qualify-lead
    type: agent
    agent: sales-qualifier

Or define a full pipeline in Workflow YAML:

astromeshctl new workflow lead-pipeline
astromeshctl run lead-pipeline --workflow --input '{"company": "Acme"}'

Visualize the DAG in VS Code with the Workflow Visualizer.

Quick Reference

Tool	What it does	Where
`astromeshctl`	Scaffold, run, debug, monitor, deploy	Terminal
Copilot	AI assistant that knows your project	Terminal + VS Code
VS Code Extension	IntelliSense, traces, metrics, workflow viz, copilot	VS Code
Dashboard	Web UI for metrics and traces	`/v1/dashboard/`