ai-project-command-skills

Use when the user writes /zno-change or asks to change existing behavior or scope in an ongoing project. Identifies the affected docs and code before implementing, updates them together with the change, and records the change in the feature changelog.

Use when the user writes /zno-continue or starts a new session wanting to resume where a previous conversation left off. Reads the status snapshot and supporting docs, summarizes the project state, then continues with the next safe task.

Use when the user writes /zno-deploy or asks about deployment, release, environments, migration, or rollback. Plans and executes the deployment change, updates docs/operations/08-deployment.md, and updates local-development docs when commands or env vars change.

Use when the user writes /zno-evaluate or wants to validate whether a project idea is worth doing before committing. Uses Socratic questioning to challenge the idea across pain points, target users, existing solutions, ROI, risks, and timeline. Outputs a structured evaluation conclusion (do / don't do / defer). Should be used before /zno-init.

Use when the user writes /zno-feature or asks to add a new feature to an ongoing project. Clarifies only what affects scope or design, follows TDD when a test framework exists, updates the relevant docs and the feature changelog, and passes the verification gate before completion.

Use when the user writes /zno-fix or reports a bug or defect. Reproduces or reasons about the defect, locates the root cause, fixes it with a regression test when possible, verifies the fix, and updates the changelog if behavior changed.

Use when the user writes /zno-goal (optionally with --super), gives a whole project or phase objective, or says they don't want to issue per-feature commands. Executes the document-first workflow autonomously through phases and tasks, stopping only at required confirmation gates. With --super, ordinary confirmations are skipped and decisions are logged instead.

Use when the user writes /zno-handoff or wants to prepare the project for another developer or AI agent to take over. Writes or updates docs/handoff/05-handoff-guide.md with entry points, common tasks, and pitfalls.

Use when the user writes /zno-init or wants to start a new software project from a requirement description. Clarifies the goal, scaffolds AI_DEVELOPMENT_RULES.md and core project docs, captures the frontend design direction and design tokens, and waits for user confirmation before any implementation.

Use when the user writes /zno-plan, has no clear next step, asks what to add or improve, or wants feature recommendations. Reads the project docs and recommends 3-7 prioritized directions with value, complexity, and risks, then waits for the user to choose. Never implements directly.

Deprecated compatibility alias of /zno-init. When invoked, apply the zno-init skill to clarify requirements and scaffold project kickoff docs.

Use when the user writes /zno-retro or wants to conduct a retrospective after completing a project phase or milestone. Compares original plan vs actual outcome, identifies lessons learned, extracts reusable rules, and writes findings to docs/development/16-retrospective.md. Generalizable lessons are appended to AI_DEVELOPMENT_RULES.md.

Use when the user writes /zno-review or wants to review code before merging. Audits across 5 dimensions - security, performance, error handling, SOLID/DRY, and project conventions - and outputs findings as blocking items, suggestions, and reminders.

Use when the user writes /zno-roadmap or wants to adjust project phases, priorities, MVP boundary, or milestones. Updates docs/product/06-roadmap.md and keeps it consistent with the current status and changelog.

Use when the user writes /zno-status, wants to save current progress before switching or ending a session, or asks for a progress snapshot. Inspects the workspace and writes a factual, handoff-ready snapshot to docs/development/10-current-status.md. Does not implement new work.

Deprecated compatibility alias. When the user writes /zno-super, apply the zno-goal skill in --super (high-autonomy) mode.

Use when the user writes /zno-tech or the work involves a major technical decision - frameworks, databases, cache, queues, auth, storage, payment, deployment, CI/CD, monitoring, or any heavy dependency. Presents 2-3 compared options, waits for user confirmation, records the decision in docs/engineering/04-tech-decisions.md, then implements.

Use when the user writes /zno-upgrade, asks to update the zno-* skill package to the latest GitHub version, or wants to add missing template docs to a project initialized by an older skill version. Runs the matching upgrade script and records upgrade history.