skills

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

Builds and sharpens a project's domain model in `.agents` docs. Use when the user wants to pin down domain terminology, record an ADR, or another skill needs to maintain project language.

A relentless interview to sharpen a plan or design. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

A relentless interview that turns a plan or design into a user-requested artifact. Use when the user wants to stress-test a direction and explicitly capture the result in project documentation.

Interview the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, uses a grill trigger phrase, or another skill needs a grilling session.

Writes a compact continuation handoff for another agent or future session. Use when the user asks to hand off, compact context, continue later, or preserve next steps.

Implement a piece of work based on a spec or set of tickets.

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.

Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in a local file, or native blocking links on a real tracker.

Plan a huge chunk of work — more than one agent session can hold — as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.

Routes ViteHub project names and aliases to their local repository roots. Use when the user mentions a ViteHub project, Quiver Agents, Quiver Babysitter, ViteHub Babysitter, Nuxt Agent, Bitacora Audio, or ViteHub Summary.

Converges Linux agent machines into private shared coding nodes. Use when bootstrapping or reconciling a VPS, Tailscale access, Docker or Portainer, Codex and GitHub auth profiles, shared workspaces, scheduled jobs, or machine cleanup.

Persists rigorous internal or external evidence for one decision. Use when design, planning, retrospectives, or architecture depends on sourced research before deciding.

Prunes one user-accepted branch inside a grilling session. Use when the user asks to skip it or accepts the current recommendation.

Drafts human-centered messages and recommends interpersonal next actions using practical relationship principles. Use when the user asks for emails, DMs, apologies, conflict repair, influence, persuasion, feedback, favors, boundaries, or advice for a personal or professional situation with another person.

Refines one GitHub PR toward review or merge readiness. Use to check a PR, address review comments, inspect CI, reduce scope, or identify remaining blockers.

Finds accidental complexity and returns the smallest useful simplification plan. Use when the user says simplify, says scope feels too complex, or wants a smaller PR, API, or direction.

Reviews code for maintainability, structure, abstractions, large files, and cleanup opportunities. Use when the user asks for strict, harsh, deep, or ambitious code review.

Root router for UI work that selects the smallest useful design, reference, implementation, or QA context. Use when the user mentions UI, ui-skills, docs.ui.sh, ui.sh, Mobbin, visual quality, spacing, color, compactness, layout, tables, selectors, buttons, responsiveness, or hierarchy.

Validates a direction on explicit request or immediately before an authorized direction hardens into a plan, durable artifact, API, workflow, merge strategy, or implementation.

Coordinates autonomous work by splitting tasks into owned sessions and keeping one owner per branch, PR, file area, and side effect. Use when the user asks for workflow, orchestration, heartbeat, child threads, background sessions, multi-agent coordination, PR readiness, or automatic merge coordination.

Cleans Git worktrees, stale branches, and build artifacts without losing unique work. Use to clean worktrees, prune branches, reduce disk usage, or remove stale Codex workspaces.

Creates or revises one portable agent skill with concise routing, progressive disclosure, and clear safety boundaries. Use when the user wants to write, clean up, or restructure a skill.

Converges authored pull requests by repairing actionable review feedback, failed checks, and branch blockers; substitutes exact-head Sol review when Codex is unavailable; and applies repository-specific ready or merge gates. Use for an autonomous PR heartbeat or review-fix loop.

Reviews reusable package craft: exports, layout, naming, tests, docs, compatibility, and release shape. Use for npm packages, SDKs, framework utilities, reusable modules, or public APIs.

Finds Codex session coordinates by id, title, repo path, rollout path, or topic. Use when a user or another skill needs thread metadata and rollout paths for its own review.

Writes or reviews GitHub pull request bodies in the repository's preferred style with automatic issue linking. Use when opening, updating, or checking a PR body.