gstack cheat sheet

Six forcing questions that reframe your product. Pushes back on your framing, challenges premises, and writes a design doc that feeds every downstream skill.

One command, fully reviewed plan. Runs CEO -> design -> eng -> DX automatically, auto-detects which apply, and surfaces only the taste decisions for your approval.

Rethink the problem. Find the 10-star product hiding inside the request. Four modes: Expansion, Selective Expansion, Hold Scope, Reduction.

Lock architecture, data flow, diagrams, edge cases, and tests. Forces hidden assumptions into the open.

Rates each design dimension 0-10, shows what a 10 looks like, then edits the plan to get there. AI-slop detection. One question per design choice.

Interactive DX review: personas, competitor TTHW benchmarks, your magical moment, friction tracing. Three modes, 20-45 forcing questions.

Build a complete design system from scratch. Researches the landscape, proposes creative risks, generates realistic product mockups, writes DESIGN.md.

Turn vague intent into a precise, executable spec in five phases with a Codex quality gate and secret redaction. --execute spawns claude -p in a fresh worktree.

Generates 4-6 AI mockup variants, opens a comparison board in your browser, collects feedback, and iterates. Taste memory learns what you pick.

Turn a mockup into production HTML/CSS with Pretext computed layout: text reflows, heights adjust. 30KB, zero deps. Detects React/Svelte/Vue.

Find the bugs that pass CI but blow up in production. Auto-fixes the obvious ones, flags completeness gaps.

Independent review from OpenAI Codex CLI. Three modes: pass/fail gate, adversarial challenge, open consultation. Cross-model analysis when paired with /review.

OWASP Top 10 + STRIDE threat model. Zero-noise: 8/10+ confidence gate, independent verification, and a concrete exploit scenario per finding.

Systematic root-cause debugging. Iron Law: no fixes without investigation. Traces data flow, tests hypotheses, stops after 3 failed fixes.

Test your app, find bugs, fix them with atomic commits, re-verify. Auto-generates a regression test for every fix.

Same methodology as /qa but report only. A pure bug report with no code changes.

The /plan-design-review audit, then fixes what it finds. Atomic commits, before/after screenshots.

Live onboarding audit: navigates docs, tries the getting-started flow, times TTHW, screenshots errors, and compares against your plan scores.

Give the agent eyes: real Chromium, real clicks, real screenshots, ~100ms per command.

Baseline page load times, Core Web Vitals, and resource sizes. Compare before/after on every PR.

Share your browser with any AI agent (OpenClaw, Hermes, Codex, Cursor). Scoped tokens, tab isolation, rate limiting, activity attribution.

Sync main, run tests, audit coverage, push, open the PR. Bootstraps a test framework if you do not have one.

Merge the PR, wait for CI and deploy, verify production health. One command from "approved" to "verified in production."

Post-deploy monitoring loop. Watches for console errors, performance regressions, and page failures.

Update every doc to match what you shipped. Builds a Diataxis coverage map so gaps show in the PR. Auto-invoked by /ship.

Generate missing docs from scratch with the Diataxis framework after reading the codebase.

Team-aware weekly retro: per-person breakdowns, shipping streaks, test-health trends. /retro global spans all projects and AI tools.

Review, search, prune, and export what gstack learned. Patterns and preferences compound across sessions.

Warns before destructive commands (rm -rf, DROP TABLE, force-push). Say "be careful" to activate. Override any warning.

Restrict file edits to one directory so the agent cannot touch unrelated code while debugging.

/careful + /freeze in one command. Maximum safety for prod work.

Remove the /freeze boundary.

Launch AI-controlled Chromium with sidebar, anti-bot stealth, and auto model routing (Sonnet for actions, Opus for analysis).

Launch GStack Browser — Chromium with the sidebar extension baked in.

Import cookies from your real browser (Chrome, Arc, Brave, Edge) into the headless session to test authenticated pages.

Zero to running gbrain in under 5 minutes. PGLite local, an existing Supabase URL, or auto-provision a new Supabase project.

Re-index this repo into gbrain and refresh the search-guidance block in CLAUDE.md. Idempotent; safe to re-run.

Drive a real iPhone over USB CoreDevice. Capability-tier allowlist, per-device session lock, audit log. --tailnet exposes it to remote agents.

Autonomous iOS bug fixer.

Designer's-eye HIG audit on real hardware.

Remove the DebugBridge SPM package and all #if DEBUG wiring.

Regenerate the iOS debug bridge against the latest upstream templates.

Upgrade gstack to latest. Detects global vs vendored install, syncs both, shows what changed.

One-time setup for /land-and-deploy. Detects your platform, production URL, and deploy commands.