nanostack

Nanostack adds planning, code review, security checks, and testing to your AI coding workflow. It saves the plan and results as files, so you can see what changed and how it was checked.

Verified adapters: Claude Code, Cursor, OpenAI Codex, OpenCode, and Gemini CLI. Claude Code supports blocking hooks; the other adapters use guided instructions. See the per-agent capabilities and verification records for dates and test scope. Installation checks do not guarantee a full sprint on every host version.

Quick start

npx create-nanostack

Run /nano-run in your agent to configure the project. See requirements and installation options.

Start with /think to work through an idea, or /nano to plan a defined change. For an autonomous sprint in an existing project, use /feature <goal>; it automatically approves the plan and continues through the phases.

What changes after installing Nanostack

• Plan the change. Record the scope, files, risks, and checks before implementation.
• Review the result. Check the code against the plan, audit security, and test behavior.
• Inspect the work. Read saved reports or open them as visual artifacts in your browser.
• Adapt the workflow. Edit the skills or add your own phases with custom stacks.

Inspired by gstack from Garry Tan. Apache 2.0. No Nanostack cloud. See privacy for telemetry and agent-provider details, and release notes for recent changes.

Try it safely first

Not sure yet? Start with a disposable sandbox from the Examples Library. It gives you a real sprint without risking your product.

Each example has a copy-paste prompt, expected sprint flow, success criteria, and reset steps. Full Examples Library: examples/.

compliance-release is advanced. It is not a starter app and it is not a compliance certification. It shows how several custom skills can compose into one release workflow.

Choose your path

What is Nanostack?

Your agent can already edit files and run commands. Nanostack gives it a method: 13 built-in skills, a seven-phase default sprint, and a framework for composing your own workflow stacks.

The default sprint turns a request into a scoped, reviewed, security-checked, tested change with a PR and a sprint journal. Each phase writes a structured artifact: a small local file that records what was decided and what was checked. Later phases read those files instead of depending only on chat history.

The bet is artifact-first delivery:

• Skills are plain text.
• Artifacts are local JSON.
• Gates verify evidence before release.
• Visual artifacts render the same evidence as local HTML.
• Custom workflow stacks extend the same pipeline with your own phases.

On Claude Code, Nanostack can enforce parts of the workflow through PreToolUse hooks. On other agents, the same workflow runs as guided instructions. See What enforces on which agent for the honest per-host table.

The built-in sprint is the default stack:

Two profiles, same rigor

Nanostack adapts the explanation, not the standard.

Local mode uses Guided language by default. A git project can still use Guided if the user wants simpler explanations.

The wording rules live in reference/plain-language-contract.md. The session fields that select the profile live in reference/session-state-contract.md.

What is enforced depends on your agent

Nanostack is agent-agnostic, but agent hosts do not expose the same control points. The adapter files in adapters/ are the source of truth for each host.

A detailed per-host matrix (Bash guard, Write/Edit guard, phase gate) lives further down in What enforces on which agent.

The sprint

Nanostack is a process, not a collection of tools. The skills run in the order a sprint runs:

/think → /nano → build → /review → /security → /qa → /ship

Each skill feeds into the next. /nano writes an artifact that /review reads for scope drift detection. /review catches conflicts with /security findings. /ship verifies everything is clean before creating the PR. On Claude Code the phase gate enforces the pipeline at the hook layer: git commit is blocked until review, security, and qa have fresh artifacts. On agents without hook support the same gate runs as guided instructions, so the safety depends on the agent following them; see What enforces on which agent.

Power tools