Optimize

wendy project optimize statically analyzes your project's build configuration — its Dockerfile(s), requirements.txt, and wendy.json — and reports missed build-speed and runtime optimizations.

It runs locally and is read-only by default. It works on a single Dockerfile, a multi-service / Compose project (findings are grouped by service), and native Swift (Package.swift) or Brewfile projects.

What it checks

• Build caches — compiled-language build/install steps (cargo, go, swift, npm/yarn/pnpm, pip) that run without a BuildKit --mount=type=cache, and so re-download or re-compile dependencies on every build.
• Release vs. debug — debug builds shipped to production (swift build without -c release, cargo build without --release), and whether WENDY_DEBUG is wired to toggle the optimization level.
• CUDA / ML — a CPU-only ML wheel (e.g. torch==…+cpu) paired with the gpu entitlement (or a CUDA wheel without it), and x86 nvidia/cuda base images on an arm64 (Jetson) target.
• Architecture & image — an amd64 base image on an arm64 device (which runs under slow QEMU emulation or fails), a missing .dockerignore, and single-stage builds that ship their full build toolchain.

Usage

wendy project optimize            # report findings (colorized in a terminal, JSON in CI)
wendy project optimize --json     # machine-readable findings
wendy project optimize --fix      # apply the safe, deterministic fixes
wendy project optimize --agentic  # emit a context bundle for an AI agent

Flags

• --fix — apply the safe fixes only: add a build-cache mount, add the release flag (swift/cargo), and create a default .dockerignore. Fixes are idempotent; contextual changes (multi-stage refactors, choosing the right CUDA wheel) are left to you or the --agentic flow.
• --agentic — instead of a report, emit a JSON bundle (static findings plus the verbatim project files and a prompt) designed to be piped into Claude Code or the Wendy MCP server.
• --severity <info|warning|error> — the minimum severity that causes a non-zero exit. Defaults to warning.
• --arch <arch> — override the target architecture (defaults to arm64).
• --json — emit findings as JSON (also the default when output is not a terminal, e.g. in CI).

Exit codes

• 0 — no findings at or above the severity threshold.
• 1 — findings at or above the threshold (use this to gate CI).
• 2 — execution error (no project found, parse failure).

At build time

After a slow incremental build (one that reused cached layers and still took more than ~50s), wendy run / wendy build will run this scan automatically in an interactive terminal, show the findings, and offer to apply the safe fixes for your next build. This never runs in CI or non-interactive shells.

Sample projects

The repository ships small, deliberately un-optimized sample projects under Examples/project-optimize-samples/ that each trigger a specific finding, so you can see the analyzer and --fix in action:

Run wendy project optimize (or --fix) inside any sample directory to reproduce the corresponding finding. See the samples README for details.

A note on --agentic and secrets

The --agentic bundle includes the verbatim contents of your Dockerfile(s), requirements.txt, and wendy.json so the agent has full context. These files can contain secrets (ARG/ENV tokens, private registry URLs). The command prints a reminder to stderr; review the bundle before sending it to an external agent.