artur-plugins

Plugin artur-plugins from marketplace artur-plugins in garbageek/claude-plugins. Commands read from the repo's manifest — yesterday.

Skills (30)

• architect

  Design software architecture SPECs, product briefs and lightweight PRDs, create/update/review ADRs, review or update designs, create, review, or update implementation plans including source-grounded change plans for existing code, compare system-design trade-offs, and carry out authorized architectu

• audit-discovery

  Use when findings should be recorded as structured audit tickets. Cold-start friendly: initialize the audit workflow, inspect evidence, then create DRAFT/OPEN tickets. Do not resolve or verify tickets.

• audit-init

  Initialize the audit ticket workspace when the user explicitly requests audit lifecycle setup. Preserve existing records and report health and queues. Not a prerequisite for read-only code review.

• audit-next

  Find the next eligible ticket for resolution or independent verification in an existing audit workspace. Use for queue selection without changing state.

• audit-resolution

  Use when the user wants to implement fixes for audit tickets. Default to one unblocked ticket from audit next. Resolution records evidence and stops at READY_FOR_VERIFICATION; it never writes PASS.

• audit-status

  Inspect existing audit health, summary, and resolution or verification queues. Use for status questions; do not initialize or change audit state.

• audit-triage

  Use when existing audit tickets need prioritization, dependency ordering, shared-file or shared-cause fix groups, or a roadmap. Do not create tickets or edit implementation code.

• audit-verification

  Use when resolved audit tickets need independent verification. Default to the next READY_FOR_VERIFICATION ticket. Verification may write PASS/PARTIAL/FAIL with criterion evidence; it should not edit implementation code.

• code-to-prd

  Reverse-engineer a frontend, backend, or mixed codebase into an evidence-backed Product Requirements Document. Use when documenting existing routes, user flows, APIs, data contracts, permissions, background work, and observable behavior without inventing missing product intent.

• codex-ubuntu-server

  Install, authenticate, update, diagnose, or recover Codex CLI on Ubuntu Server, including managed remote control and independent self-maintenance handoff. Use for headless Ubuntu hosts; not for desktop app updates or general server administration.

• copilot-review-customizer

  Research, design, and ship repository-specific GitHub Copilot Code Review customization for a given repository. Use when the user wants Copilot PR reviews tailored to a project, fewer speculative or generic Copilot findings, or asks to create `.github/skills/code-review/SKILL.md`, review-focused Cop

• create-plugin

  Use when the user requests a new Claude Code or Codex plugin, a reusable workflow packaged for either or both hosts, or creation of a private plugin through the connected Plugin Creator backend.

• create-skill

  Create, edit, evaluate, or optimize a reusable Agent Skill (SKILL.md) for Claude Code, Codex, or both. Use for standalone skills, skill instructions inside an existing plugin, skill trigger descriptions, evaluations, benchmarks, and .skill packaging. Route whole-plugin creation and release managemen

• debug

  Investigate a bug, failing command, build or CI failure, crash, wrong output, performance regression, or other unexpected behavior to its root cause, and fix it when a fix is requested. Use when something is broken and the cause is not yet established, or when earlier fix attempts did not resolve it

• deep-review

  Review code, diffs, regressions, and implementation claims using traceable execution paths. Separate confirmed defects, risks, and improvement suggestions. Use specialist lenses for UI click paths, existing test quality, and operator-facing truth surfaces. Review is read-only; ticket creation and re

• design-prompts

  Create, rewrite, debug, adapt, and evaluate prompts for language models and AI agents. Use when the user asks for a copy-ready prompt, prompt optimization, a system or developer prompt, structured-output instructions, few-shot examples, prompt migration between tools or models, prompt diagnostics, a

• diagnose-session

  Investigate what went wrong in a recorded Claude Code or Codex agent session, such as a skill that did not trigger or triggered wrongly, an ignored plan, repeated work after compaction, stumbles, unsupported completion claims, or unexpected cost or duration, by reading its transcripts on disk and re

• feature-scattering

  Investigate feature locality, shotgun-change costs, divergent change, and responsibility boundaries using source distribution and optional Git co-change signals. Use when changing one feature requires coordinated edits across unrelated modules. Read-only analysis by default; confirmed findings use t

• human-execution-runbook

  Draft, update from execution evidence, or resume a sequential runbook of actions a human personally performs. Use for manual approvals, external UI steps, release/deployment procedures, operational observation, physical actions, and interruption/retry handling. Keep human authority, attempts, observ

• humanizer

  Drafts, rewrites, and checks text that goes out under the user's own name (reports, status updates, design docs, proposals, emails, Slack messages, meeting notes, talking points for a manager or leadership) so it reads as written by that person and carries no AI tells. Use whenever drafting or editi

• init-repo-instructions

  Create repository-specific agent instructions for the requested Claude or Codex scope from real commands, boundaries, and non-obvious constraints. Preserve existing guidance; use when initializing instructions, not transforming the whole repository.

• inspect-plugin

  Use for read-only inspection or per-skill quality assessment of a Claude Code or Codex plugin from source, an archive, or an accessible Plugin Creator release, including structure, workflow boundaries, compatibility, integrations, and release history.

• install-agent-plugins

  Use when installing, enabling, disabling, upgrading, removing, or troubleshooting installed plugins and marketplaces in Claude Code, the Claude app, or Codex. Also use for local plugin development loading and translating installation commands between hosts; not for editing plugin source or publishin

• iphone13mini-ui

  Build or fix web UI for iPhone 13 mini Safari: 375px layouts, overflow, touch controls, safe areas, toolbar or keyboard overlap, and screenshot comparison. Use for small-iPhone web UI work, not native iOS apps or general desktop styling.

• local-wiki

  Build or update a source-backed local repository wiki, validate its navigation, or export a linked Markdown bundle and optional offline static site. Use for implemented-system documentation, not future design or automatic documentation cleanup.

• recover

  Assess and recover an existing project after suspect changes caused regressions, incomplete replacements, authority splits, compatibility loss, or scope drift. Compare trusted baselines and preserve valid improvements. Assessment is read-only by default; execute only an explicitly authorized recover

• review-feedback

  Use when assessing or addressing specific pull request review comments or supplied code-review feedback; separate evidence-backed judgments from authorized code changes and GitHub thread actions.

• review-repo-instructions

  Review scoped repository instructions against actual code, commands, and Claude/Codex loading behavior. Identify drift, conflicts, noise, and useful retained rules; return source-backed minimal changes without editing unless explicitly requested.

• uncompromising-handover

  Write a fidelity-first context handover (context-checkpoint.md) so a different agent, model, harness, or fresh session can continue an in-progress coding task with zero prior context. Use when the user asks to hand off, hand over, checkpoint, compact, or save session context, or before switching mod

• update-plugin

  Use when changing plugin source or publishing a revised Claude Code/Codex package, including cross-platform conversion, skill/file removal, local/archive edits, or an eligible Plugin Creator release. Not for upgrading or removing an installed client copy.

Subagents (5)

• audit-discovery

  Evidence-backed audit ticket discovery. Use to inspect code/logs, initialize the audit workflow, and create DRAFT/OPEN tickets only from concrete findings.

• audit-resolution

  Implements fixes for selected unblocked audit tickets and records resolution evidence. Stops at READY_FOR_VERIFICATION.

• audit-triage

  Prioritizes existing audit tickets, records executable triage metadata, and models dependencies consumed by audit next.

• audit-verification

  Independently verifies READY_FOR_VERIFICATION audit tickets and writes PASS/PARTIAL/FAIL verdicts through audit verify.

• deep-code-auditor

  Use for deep, read-only code and diff audits, bug-fix validation, regression checks, and correctness investigations. Trace related execution paths and report evidence-backed defects, risks, and coverage limits without modifying files.

Similar plugins