Everform CLI

Your AI workflows are running your business. Eve tells you which ones are working, which are failing silently, and makes them better over time.

Quick Start

# Install globally
npm install -g @everform/cli

# Or run without installing
npx @everform/cli process test my-workflow.yaml

# Or use Bun
bun add -g @everform/cli

Configure your LLM provider

# Set your provider (Anthropic is default)
eve config set LLM_PROVIDER anthropic
eve config set ANTHROPIC_API_KEY sk-ant-...

# Or use free local models — no API key needed
eve config set LLM_PROVIDER local
ollama pull llama3 && ollama serve

Commands

Design processes from natural language

# Convert an SOP document to a process DAG
eve ingest onboarding-sop.md

# Validate the generated process
eve process validate onboarding-sop.yaml

Score and test

# Structural analysis (no LLM needed)
eve process score onboarding-sop.yaml

# The killer feature: design-time validation loop
# CLI tests → deterministic scoring → Eve improves → repeat → converge
eve process test onboarding-sop.yaml --max-iterations 10 --budget $5

Run and export

# Execute with governance gate + real-time progress
eve process run onboarding-sop.yaml

# Export to governance.yaml + SKILL.md
eve process export onboarding-sop.yaml

System overview

# See your organization as a constellation
eve system inspect

CI Integration

# JSON output for pipelines
eve process test workflow.yaml --json | jq '.best_score'

# Exit code: 0 = converged, 1 = not converged
eve process test workflow.yaml --convergence-threshold 85
echo $?

What makes Eve different

1. Automation suitability scoring — real-time, automatic, from execution data
2. Paradigm prescription — Eve recommends deterministic/AI-driven/hybrid per step
3. Organizational learning — the expertise ratchet that never loses knowledge

LLM Providers

| Provider | Setup | Cost | |----------|-------|------| | Anthropic (default) | eve config set ANTHROPIC_API_KEY sk-ant-... | Subscription or pay-per-use | | OpenAI | eve config set LLM_PROVIDER openai | Subscription or pay-per-use | | OpenRouter | eve config set LLM_PROVIDER openrouter | Pay-per-use (100+ models) | | Local (Ollama) | eve config set LLM_PROVIDER local | Free |

Process DAG Format

Processes are defined as YAML DAGs:

name: customer-onboarding
description: Onboard new customer with welcome email and account setup
nodes:
  - id: collect-info
    prompt: "Extract customer name, email, and plan from the signup form data"
    paradigm: dag_node

  - id: create-account
    bash: "curl -X POST https://api.example.com/accounts -d '...'"
    paradigm: dag_node
    depends_on: [collect-info]

  - id: send-welcome
    prompt: "Write a personalized welcome email for the new customer"
    paradigm: dag_node
    depends_on: [create-account]

Links

• GitHub
• Everform — the organizational OS for AI
• Report a bug