latentwill

Orient an agent USING xvision as a product — running `xvn` CLI verbs, the embedded dashboard, MCP tools, strategy/eval workflows, the live Tailscale nodes, and the remote CLI job surface. USAGE ONLY. If the task is editing crates/**, frontend/web/**, migrations, contracts, or CI/deploy scripts, use `xvision-dev` instead; do not load this skill just because the cwd is the xvision repo.

Use when QAing the xvision app through direct API or CLI calls, especially to verify Strategy, Scenario, and Eval create/edit/delete behavior, detect manifest drift, duplicate records, or invalid model resolution, and collect raw HTTP evidence without the browser UI.

Use when asked for a UX/UI audit, design review of the running product, "audit the site", or before a demo/launch — boots the real site, screenshots every page (desktop + mobile), audits like a senior design lead on day one, prioritizes P0-P3, fixes safe small issues on the spot, and writes an actionable report to docs/design-audit/.

Operate the xvision autooptimizer: distill Observations into candidate Patterns, run the gate against today and the untouched test period, record blind findings, activate or retire Patterns, and preserve the lineage evidence for audit.

Operate and audit the xvision self-improvement flywheel across memory capture, autooptimizer, training run example pools, child training, lineage, velocity, and look-ahead-protection evidence.

This skill is for interface design — dashboards, admin panels, apps, tools, and interactive products. NOT for marketing design (landing pages, marketing sites, campaigns).

Operate xvision's persistent memory layer: inspect Observations and Patterns, preview recall safely, manage Pattern lifecycle, run leakage probes, and avoid F+L+T violations.

Use when starting a new team-coordination track in the xvision repo. Walks through the sync-before-work ritual — checks `team/board.md`, validates the worktree state against `team/contracts/<slug>.md`, runs `scripts/board-lint.sh`, and outputs the briefing statement before any code is touched. Pair with `xvision-dev`.

Use when expanding, varying, inverting, recombining, or systematically exploring an idea — brainstorming alternatives, generating opposites, finding unexpected angles, breaking out of a creative rut, or producing a fresh framework instead of the same generic approach. Renders the expansion as plain-text-portable output that survives any channel (Telegram, SMS, plain Slack, fixed-width-only email). Each invocation produces a different combination of ideational operations so the agent can't fall into a default routine. For monospace-rich rendering with Unicode box-drawing and figlet banners, see the sibling skill `ideonomy-rich`.

Use when expanding an idea AND the output channel can render monospace cleanly — terminals, READMEs, blog posts with monospace code blocks, fixed-width-font emails, ttyrec sessions, the Claude Code transcript itself. Same primitives as `ideonomy-plain`, but renders the artifact as performative ASCII art with Unicode box-drawing, figlet banners, density gradients, and visible ideonomy-machinery layers (tuple legend, dimensions surfaced, operator-named dividers, ideonomy trail). If the channel might mangle Unicode (Telegram, SMS, plain Slack DMs), use `ideonomy-plain` instead.

Second-order thinking engine. Forces first-principles reframing, temporal consequence tracing ('and then what?'), structured evaluation, and a hard sycophancy override before delivering any non-trivial response. MUST trigger when: (1) the query is complex, multi-step, or has ambiguous intent; (2) the user seems confused, frustrated, or stuck; (3) you're about to give advice that has real consequences (architecture, strategy, financial, interpersonal); (4) the problem involves tradeoffs, unknowns, or competing concerns; (5) you're answering from pattern-matching rather than actual reasoning; (6) the user says 'think harder', 'think about this more', 'are you sure', 'that doesn't seem right', '/vigil'; (7) you're generating code, configs, or recommendations that will be deployed. Applies to ALL domains: code, strategy, writing, research, comms, debugging, planning. Do NOT trigger on trivial queries (greetings, simple factual lookups, small talk) or when the answer is genuinely obvious and low-stakes.

Deploy and manage GPU/CPU pods, network volumes, and templates on RunPod. Use when the user asks to launch a GPU server, manage cloud compute, run ML training or inference on remote GPUs, or interact with the RunPod platform in any way. Triggers: runpod, gpu cloud, launch pod, spin up server, deploy gpu, remote training, cloud gpu, inference server.