xkazm04

Session-after-session product perfection loop. The strongest available model (Fable) directs — it walks the repo's context map context-by-context, proposes 5 challenged, high-value directions per context (features, design elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Generate images with OpenAI gpt-image-2 (primary) or Leonardo AI (fallback), remove backgrounds, analyze with Gemini vision, and write SVG. For brand assets, UI illustrations, backgrounds, and icons.

Upgrade a generic UI icon or loading/empty state into a traced, motion-animated SVG. Generates flat trace-friendly art (via /leonardo tools), validates with Qwen vision, vectorizes to a clean multi-path SVG, and emits a Motion (framer-motion) reveal component. For icon + loading-state visual upgrades — NOT raw image generation (use /leonardo for that).

Hunts the highest-value surface of an LLM-powered app — the LLM call sites themselves — and drives them to their potential across three lenses. (1) Code quality of the AI plumbing (wrapping/chokepoint, logging/telemetry, caching/dedupe, schema+validation+self-repair). (2) Business value, by reusing the UAT Character method (representative users with jobs-to-be-done + a senior-quality bar + time-saved) but TESTING ONLY THE LLM PIECES — does each prompt's grounding and output clear the bar. (3) Model optimization as an alternative scenario — benchmark the same character inputs across models × thinking levels to find quality degradation/upgrade vs cost/latency. Everything is memorized in a linked **Obsidian vault** (one note per call site / character / model / session) so each scan builds on the last. Stack-agnostic engine; per-app specifics live in the repo's `tiger/` overlay (which IS the vault). Invoke with `/tiger init|scan|run|benchmark|recall|backlog [args]`.

Simulated User Acceptance Testing for the Personas desktop app, driven by Characters (representative users with jobs-to-be-done) rather than feature/code coverage. A capable LLM verifies each user journey in two chronological certification levels — L1 theoretical (over a code-derived surface model, cheap + mass-parallel) then L2 empirical (the LIVE app driven through the test-automation harness, serial) — judging through each Character's own consistent lens (time saved vs the manual way, and senior-in-role quality). Personas-specific: L1 reads context-map.json + the React/Rust source; L2 drives the running app via the test-automation server on :17320 (NOT a browser). Per-run specifics live in the repo's uat/ overlay. Invoke with `/uat init|update|run|promote [args]`.

Iteratively prototype a UI component through directional variants behind a tab switcher, then consolidate and refactor the winner. Use when the user wants to improve a component they consider a pillar of the app (visual appeal, creativity, UX clarity).

Personalized AI-native onboarding for xkazm04/personas — generated by Ascent from a scan on 2026-06-10 (currently L4 Integrated, 75/100). Run it to adopt the highest-leverage practices (D4, D9, D8, D2) that move this repo toward autonomous, LLM-driven development.

Production-readiness code review for Rust backend and React frontend changes

Session-after-session product perfection loop for Vibeman. The strongest available model (Fable) directs — it walks the repo's context map context-by-context, proposes 5 challenged, high-value directions per context (features, design elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Run a Vibeman pipeline on a project. Three modes — (A) goal-based development (PLAN > IMPLEMENT > VERIFY > REPORT), (B) audit-driven scan + triage + wave-based fix implementation, or (C) scan-and-decide (pick one context group; the skill auto-selects Idea scanners, generates a capped backlog, you accept/reject each, then it implements the approved scope). All run with quality gates, brain-signal recording, and structured per-wave/per-phase reporting.

Systematically triage, challenge, and process a backlog of auto-generated idea files. Groups by context area, validates against codebase, filters BS, presents one approval gate, then executes autonomously with code review.

Generate images with Leonardo AI (Lucid Origin model), remove backgrounds, and write SVG. For brand assets, UI illustrations, backgrounds, and icons.

Fetch npm package updates, analyze new features, and identify improvement opportunities for this app. Use when the user wants to explore what's new in dependencies or plan next development directions.

Review code for quality, security, and patterns. Scans changed files for bugs, OWASP vulnerabilities, SQLite injection, missing error handling, and anti-patterns specific to this Next.js + SQLite + Zustand stack.

Test-driven development workflow. Write failing tests first, then implement minimal code to pass, then refactor. Uses Vitest with better-sqlite3 test database. Enforces RED → GREEN → IMPROVE cycle.

Connect this repository to the org's Brainiac memory. Pairs the repo with the Brainiac console (device-flow approval), writes a project-scoped API key into .env without ever displaying it, verifies connectivity end to end, enriches CLAUDE.md/AGENTS.md with memory conventions, and optionally registers the MCP server and syncs the org's skill library. Safe to re-run.

Session-after-session product perfection loop. The strongest available model (Fable) directs — it walks the repo's context map context-by-context, proposes 5 challenged, high-value directions per context (features, design elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates builder subagents (model sized to the work — Sonnet for S/M briefs, Opus for L or high-risk ones) per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Simulated User Acceptance Testing for Brainiac, run as a controlled trial inside a simulated multi-stack engineering company rather than a feature walkthrough. Characters are developers (Rust/TS/Python/infra) doing real coding work; every journey runs in three arms — cold agent, Claude's native local memory (the free baseline), and Brainiac — and the verdict is the DELTA, which is allowed to come out negative. Two chronological certification levels: L1 theoretical (over a code-derived model of what an agent actually receives, cheap + mass-parallel) then L2 empirical (live server + workers + real agent sessions against the Meridian company). Ships a Harm Ledger alongside findings. Per-run specifics live in the repo's uat/ overlay. Invoke with `/uat init|update|run|promote [args]`.

Iteratively prototype a console page/component through directional variants behind a tab switcher, then consolidate the winner. Use when the user wants to explore multiple design approaches for a Brainiac console surface (visual appeal, creativity, UX clarity).

Session-after-session product perfection loop for personas-web. The strongest available model (Fable) directs — it walks the repo's context map context-by-context, proposes 5 challenged, high-value directions per context (features, design elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Generate branded app icons (favicons, PWA icons, logos) with Leonardo AI, evaluate with Gemini vision, and export a correctly-sized icon set. Optimized for clean icon SHAPE and CONTENT with minimal background noise — unlike /leonardo, which targets general scenes and illustrations.

Generate images with Leonardo AI (Lucid Origin model), remove backgrounds, analyze with Gemini vision, and write SVG. For brand assets, UI illustrations, backgrounds, and icons.

Fetch npm package updates, analyze new features, and identify improvement opportunities for this app. Use when the user wants to explore what's new in dependencies or plan next development directions.

Generate images with OpenAI gpt-image-2 (primary) or Leonardo AI (fallback), remove backgrounds, analyze with Gemini vision, and write SVG. For brand assets, UI illustrations, backgrounds, and icons.

Upgrade a generic UI icon or loading/empty state into a traced, motion-animated SVG. Generates flat trace-friendly art (via /leonardo tools), validates with Qwen vision, vectorizes to a clean multi-path SVG, and emits a Motion (framer-motion) reveal component. For icon + loading-state visual upgrades — NOT raw image generation (use /leonardo for that).

Iteratively prototype a gravitone-web UI module through 2 directional variants behind a tab switcher, then consolidate the winner. Use per module (playground, voices, API keys, auth) to pick the best UI/UX.

Iteratively prototype an ascent UI surface through directional variants behind a tab switcher, then consolidate and refactor the winner into the brand system. Use when the user wants to level up a component they consider a pillar of the app (visual appeal, creativity, UX clarity) — e.g. an org dashboard panel, a report view, the launch star-map, or an onboarding step. Not for fixed-scope tweaks or bug fixes.

Identifies missing features and enhancement opportunities. Use for a focused Feature Scout pass over a project or a diff.

Reviews user experience patterns and suggests improvements. Use for a focused UX Reviewer pass over a project or a diff.

Session-after-session product perfection loop. The strongest available model (Fable) directs — it walks the repo's context map context-by-context, proposes challenged, high-value directions per context (features, design elevations, significant optimizations), gates each round's slate with the user, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Iteratively prototype a UI surface through directional variants behind a tab switcher, then consolidate and refactor the winner. Use when the user wants to master a component/page they consider a pillar of the app (visual appeal, creativity, UX clarity).

Session-after-session product perfection loop for PoF. The strongest available model (Fable) directs — it walks context-map.json context-by-context, proposes 5 challenged, high-value directions per context (features, design elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in the Obsidian pof vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Use when the user shares a YouTube/article URL or raw text about Claude/AI game-dev automation, MCP/Unreal tooling, agent harnesses, or a dev-tool showcase and wants ideas PoF can adopt into its headless tooling or UI workflows. Triggers: '/research <url>', 'can we use this for PoF', 'research this video/technique', 'any ideas from this'.

Session-after-session product perfection loop for LightTrack. The strongest available model (Fable) directs — it walks context-map.json context-by-context, proposes 5 challenged, high-value directions per context (features, API/UX elevations, significant optimizations), gates them with the user until 10 are accepted, then orchestrates one Opus builder subagent per context in isolated worktrees while making every review/merge decision itself. All state lives in a linked Obsidian vault so any future session resumes the loop exactly where the last one stopped. Invoke with `/perfect [init|propose|build|status|reflect] [context-name]`.

Guided, chronological onboarding to install, configure, and DEPLOY LightTrack onto the user's preferred stack. Walks the user through picking a database (SQLite/DuckDB/libSQL local; Postgres / Neon / Supabase / Cloud SQL / RDS / Azure DB; Firestore; BigQuery analytical) and a deploy target (local Docker Compose; Cloud Run / App Runner / Container Apps; Kubernetes via Helm), collects the required credentials securely, then performs the integration and deployment on the user's behalf and verifies it. Use when the user wants to set up, install, configure, onboard, or deploy LightTrack — e.g. "set up LightTrack", "deploy to AWS/GCP/Azure", "run it on Postgres", "get me started".

Comprehensive code review for Next.js projects. Checks for bugs, performance issues, security vulnerabilities, and adherence to best practices.