Salesforce CLI Plugin

Dependency-aware deployment orchestration for Salesforce metadata.

Smart Deployment scans metadata, builds a dependency graph, generates staged deployment waves, validates project state, and supports optional AI-assisted prioritization and dependency inference.

Explore Commands View Repository

Analyze

Scan a Salesforce project, build a dependency graph, detect cycles, and generate deployment waves.

sf smart-deployment analyze \
  --source-path force-app \
  --use-ai

Validate

Check deployment readiness, summarize issues, and optionally add AI-assisted risk assessment.

sf smart-deployment validate \
  --source-path force-app \
  --use-ai

Deploy

Run staged deployments, persist local state, and support conservative remediation for supported cycles.

sf smart-deployment start \
  --source-path force-app \
  --target-org myorg

What It Covers Today

Dependency graph generation and staged wave construction
Cycle detection and conservative remediation for supported ApexClass cycles
CLI flows for analyze, start, validate, status, resume, and config
JSON and HTML analysis reports
Optional AI assistance through shared provider adapters

Current AI Surface

The current provider abstraction supports agentforce and openai. AI can assist with dependency inference, priority weighting, and validation, while retaining deterministic fallbacks where possible.

{
  "llm": {
    "provider": "openai",
    "model": "gpt-4o-mini",
    "timeout": 30000
  }
}

6 Commands

analyze · start · validate

Plus status, resume, and config.

2 LLM Adapters

agentforce · openai

Shared provider interface for future expansion.

1 Goal

safer deploys

Less guesswork around dependency order, validation, and project readiness.

Start With The Command Surface

The most important operational docs are the command reference, AI configuration guide, and known limitations. They reflect the current state of the product better than the historical design docs.

CLI Reference AI Configuration Known Limitations