Deploy Doctor

deploy-doctor is a command-line tool that evaluates containerized applications for deployment readiness across common platforms.

It runs static checks (Dockerfile, build context, env/DB heuristics), optional runtime probes, profile-aware thresholds, and outputs machine-readable summaries for CI.

Install

Homebrew (recommended)

brew tap navig-me/deploy-doctor
brew install deploy-doctor

Curl installer

curl -fsSL https://raw.githubusercontent.com/navig-me/deploy-doctor/main/scripts/install.sh | bash

Curl installer (pinned version)

VERSION=v0.1.0 curl -fsSL https://raw.githubusercontent.com/navig-me/deploy-doctor/main/scripts/install.sh | bash

Build from source

make build
./bin/deploy-doctor version

Command List

• deploy-doctor scan: run deployment checks in the current repository.
• deploy-doctor doctor: verify local prerequisites (for example Docker availability).
• deploy-doctor profiles list: list supported deployment profiles.
• deploy-doctor profiles list --recommended: show profile recommendations from repository signals.
• deploy-doctor profiles explain <profile>: print profile details and thresholds.
• deploy-doctor version: print installed CLI version.

How scan Works

scan executes in phases:

1. Determine profile (--profile or --auto-profile).
2. Run static checks:
   • Dockerfile checks
   • context/image metadata checks
   • env/DB heuristics
   • provider config signals (evidence-driven confidence)
3. Optionally run runtime probes (unless --static-only).
4. Aggregate findings and compute:
   • severity counts
   • score
   • status
5. Emit:
   • human-readable summary
   • machine-readable CI line (DD_SUMMARY ...)

scan Flags

• --profile <name>: force profile (default: generic).
• --auto-profile: detect likely provider profile from repo files.
• --static-only: skip runtime probes.
• --runtime=false: disable runtime probes explicitly.
• --fail-on <none|critical|warning|suggestion>: fail by severity policy.
• --strict: treat warnings as failures.
• --verbose / -v: print detailed execution and finding evidence.

Output and Exit Behavior

• Human-readable summary line:
  • Scan summary: score=... status=... profile=... findings=(...)
• Machine-readable CI summary line:
  • DD_SUMMARY score=... status=... critical=... warning=... suggestion=... fail=... profile=...

Exit semantics:

• 0: scan completed without failure policy trigger.
• 1: findings violated --fail-on or --strict policy.
• 2: tool/runtime error.

Profiles

Examples:

• generic
• lightsail
• render
• railway
• flyio
• ecs-fargate
• digitalocean-app-platform
• gcp-cloud-run
• azure-container-apps
• dokku
• vps-systemd-docker

Use:

deploy-doctor profiles list
deploy-doctor profiles explain lightsail
deploy-doctor scan --profile lightsail --static-only

Quickstart

deploy-doctor doctor
deploy-doctor scan --static-only
deploy-doctor scan --auto-profile -v

CI Usage Example

deploy-doctor scan --fail-on warning --static-only

Parse DD_SUMMARY in CI logs for machine-readable gating or metrics.

Documentation

• First Scan in 3 Minutes
• Rules Catalog
• Profile Selection Guide
• Ignore/Suppression Best Practices
• Troubleshooting Playbook
• Release Playbook