chaos-land
01Command
Approve & land chaos work — move validating tickets to complete, then merge their chaos/ branches to main, autonomously resolving any conflicts.
Command
Approve & land chaos work — move validating tickets to complete, then merge their chaos/ branches to main, autonomously resolving any conflicts.
Command
Chaos-mode worker — drive one ticket autonomously to validating (used by the chaos supervisor; also runnable by hand).
Command
Enable or manage the optional Firestore mirror of the ticket board (status → Firestore via local Google credentials).
Command
Fork this Claude conversation into a new skua dashboard terminal — a divergent copy carrying the full history up to now, in the same directory. Optional label, e.g. /fork chase the upstream API error.
Command
Complete a security review of the pending changes on the current branch.
Command
Run the deterministic headless-browser smoke check on the project's web app — fails on console errors / uncaught exceptions / stuck spinners.
Skill Claude CodeCodex
Accessibility audit that files [a11y, ai-proposed] tickets for WCAG 2.2 AA gaps. ENGINE mode wraps the real tools (axe-core / pa11y) against a running dev server for objective, high-signal findings; STATIC mode (the autonomous fallback) reads components/templates for high-confidence issues (missing alt text, unlabeled…
Skill Claude CodeCodex
Owns the .tickets/ADRs/ lifecycle: create, enrich, transition (FSM-validated), link-to-ticket, list-open, promote-draft-tickets, validate-adr. User input is intentionally minimal — every create / enrich op opens with a Researcher-vs-Builder mode prompt and the AI fills in Decision, Consequences, references, and…
Skill Claude CodeCodex
External research synthesis for architectural decisions. Given a topic or named decision, WebSearches reputable architecture sources (Nygard, Fowler, AWS/GCP/Azure Architecture Frameworks, OWASP, pattern catalogs, project-internal docs), dedupes against existing ADRs, and produces a drafted ADR: Suggested position…
Skill Claude CodeCodex
Multi-agent bug hunt over the repo containing .skua/: fans out finders across correctness, error-handling, resource/state, risky-pattern, and security dimensions; adversarially refutes every candidate; files survivors as actionable 0-backlog tickets via the ticket CLI. Caps tickets per run and logs what it truncated…
Skill Claude CodeCodex
Chaos mode — skua's third, fully-autonomous ticket-execution mode (alongside user-driven and agentic). Arms a background supervisor that drains the backlog with ZERO human input: one fresh claude -p per ticket in its own chaos/TKT-NNN git worktree, self-deliberating on technical decisions via competing viewpoint…
Skill Claude CodeCodex
The generative counterpart to bug-scan: invents NEW features the user would value, grounded in the actual codebase. Fans out divergent 'product lens' agents (power-user, onboarding, adjacent-workflow, data-leverage, delight, integration), each proposing concrete features with a value hypothesis and a where-it-plugs-in…
Skill Claude CodeCodex
Optional Firestore mirror of the skua ticket board. Reflects each ticket's status + metadata into a Firestore collection so the board can be watched from OUTSIDE the repo (a phone, a shared page, a cron report). Zero-dependency: mints a token from your local Google Application Default Credentials and calls the…
Skill Claude CodeCodex
Whole-repo audit for over-engineering. Like ponytail-review, but scans the entire codebase instead of a diff: a ranked list of what to delete, simplify, or replace with stdlib/native equivalents. Use when the user says "audit this codebase", "audit for over-engineering", "what can I delete from this repo", "find…
Skill Claude CodeCodex
Harvest every ponytail: comment in the codebase into a debt ledger, so the deliberate shortcuts and deferrals ponytail leaves behind get tracked instead of rotting into "later means never". Use when the user says "ponytail debt", "/ponytail-debt", "what did ponytail defer", "list the shortcuts", "ponytail ledger", or…
Skill Claude CodeCodex
Show ponytail's measured impact as a compact scoreboard: less code, less cost, more speed, from the benchmark medians. One-shot display, not a persistent mode, and not a per-repo number. Trigger: /ponytail-gain, "ponytail gain", "what does ponytail save", "show ponytail impact", "ponytail scoreboard".
Skill Claude CodeCodex
Quick-reference card for all ponytail modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /ponytail-help, "ponytail help", "what ponytail commands", "how do I use ponytail".
Skill Claude CodeCodex
Code review focused exclusively on over-engineering. Finds what to delete: reinvented standard library, unneeded dependencies, speculative abstractions, dead flexibility. One line per finding: location, what to cut, what replaces it. Use when the user says "review for over-engineering", "what can we delete", "is this…
Skill Claude CodeCodex
Forces the laziest solution that actually works, simplest, shortest, most minimal. Channels a senior dev who has seen everything: question whether the task needs to exist at all (YAGNI), reach for the standard library before custom code, native platform features before dependencies, one line before fifty. Supports…
Skill Claude CodeCodex
Proactive, whole-repo security audit built on the stack-agnostic /security-review engine. Runs the engine over the codebase, then composes raw findings into one posture report: CVSS-anchored severity, deduplication via (CWE, normalized-resource) key, open-ticket lookup (links to existing TKT-NNN), snapshot diff vs…
Skill Claude CodeCodex
Author, audit, or refactor a Claude Code skill — produces a populated SKILL.md skeleton, validates frontmatter against the official Anthropic spec, checks against the project authoring checklist with severity-ranked output (P0/P1/P2), and audits the whole skill portfolio. Owns the canonical authoring guide…
Skill Claude CodeCodex
Cold-start bootstrap for a Claude Code skill portfolio. Introspects a target codebase (deploy units, data layers, cross-cutting concerns) and PROPOSES a customized skill portfolio + harness scaffold as a markdown plan. Child tickets via ticket-manager.create-ticket execute the scaffolding through…
Skill Claude CodeCodex
Portfolio-level skill curation: proposes merges, renames, retirements, and orchestrator wraps. Consumes skill-builder's overlap-check, audit-all, and list-orphans primitives (does NOT reimplement). Generates a versioned PORTFOLIOREORGPLAN.md and sweeps connectsto edges after structural changes.
Skill Claude CodeCodex
Knowledge + debugging playbook for the skua dashboard's in-browser terminal — the xterm.js (DOM renderer) → ttyd → zellij → shell stack under .skua/. Encodes the layer-localization diagnostic (the Playwright buffer-probe that reads xterm's buffer and compares it to ground truth, plus bypass-the-persistence-layer and…