kit

Plugin kit from marketplace ku5ic in ku5ic/claude-kit. Commands read from the repo's manifest — 2 days ago.

Skills (38)

• investigate

  Read-only investigation. Use when the user asks how, why, or where something works in the code, or reports unexpected behavior without asking for a change. Ends at findings, never edits.

• audit

  Targeted read-only audit of a code surface (a11y, debt, doc-drift, perf), or re-verify a prior audit report

• backup-patterns

  Backup patterns for the 3-2-1 rule, rsync, pg_dump, restic, retention, and restore tests. Load before any edit to a backup script or job, however small, and for backup questions.

• bash-patterns

  Bash and shell patterns for safety flags, quoting, traps, and ShellCheck. Load before any edit to a .sh, .bash, or .zsh file, shebang script, or Makefile recipe, however small, or for shell questions.

• deps

  Merge open Dependabot PRs and reconcile GitHub security alerts

• django-patterns

  Django patterns for the ORM, views, migrations, and settings. Load before any edit in a Django app (models, views, migrations, settings), however small, and for Django questions.

• docker-patterns

  Docker patterns for Dockerfiles, multi-stage builds, Compose, networking, and volumes. Load before any edit to a Dockerfile, Compose file, or .dockerignore, however small, and for container questions.

• drf-patterns

  Django REST Framework patterns for serializers, viewsets, permissions, and throttling. Load before any edit to a serializer, viewset, or API view, however small, and for DRF questions.

• engineering-fundamentals

  Design and code-level checks for modularity, size, nesting, naming, errors, and red-green-refactor. Load before planning a change and before any code edit, however small, in any language.

• fastapi-patterns

  FastAPI and Pydantic v2 patterns for Depends, response models, async, errors, and auth. Load before any edit to a file importing fastapi, however small, and for FastAPI questions.

• git-patterns

  Git patterns for Conventional Commits, rebase, reflog recovery, bisect, worktrees, and stash. Load before any commit, rebase, or other git operation, however small.

• go-patterns

  Go patterns for errors, context, goroutines, interfaces, recent release features, and tests. Load before any edit to a .go file or go.mod, however small, and for Go questions.

• graphql-patterns

  GraphQL client patterns for operations, fragments, the codegen client preset, and Apollo Client 4. Load before any edit to a .graphql or .gql file, however small, or for GraphQL questions.

• investigate

  Read-only investigation. Use when the user asks how, why, or where something works in the code, or reports unexpected behavior without asking for a change. Ends at findings, never edits.

• javascript-patterns

  JavaScript patterns for modern syntax, async, ESM and CJS, and errors. Load before any edit to a .js, .mjs, .cjs, or .jsx file or package.json, however small, and for JavaScript questions.

• kit-rules

  The kit's working rules for every reply and task. Load at the start of every session and subagent, before any other tool call.

• logging-patterns

  Python logging patterns for stdlib config, structlog, levels, handlers, and JSON logs. Load before any edit to Python logging or structlog code, however small, and for logging questions.

• meta

  Authoring and reflection. Sharpen a prompt, refresh reference skills against current releases, draft a new pattern skill pack, write a repo's conventions as cited rules, or run a retrospective

• reference
• monitoring-patterns

  Prometheus patterns for prometheus-client metric types, naming, labels, and health checks. Load before any edit to Python metrics or health-check code, however small, and for monitoring questions.

• next-app-router-patterns

  Next.js App Router patterns for the server/client boundary, caching, server actions, streaming, and runtimes. Load before any edit in app/ or next.config.*, however small, or for Next.js questions.

• nuxt-patterns

  Nuxt 3+ patterns for rendering modes, useFetch and useAsyncData, server routes, SSR state, and Nitro. Load before any edit in a Nuxt app or to nuxt.config.*, however small, and for Nuxt questions.

• opentofu-patterns

  OpenTofu and Terraform patterns for moved, import, and removed blocks, for_each, state secrets, and the lock file. Load before any edit to a .tf or .tfvars file, however small, and for HCL questions.

• pr

  Draft what lands in a PR thread (your PR's title and body, a review of someone else's PR, replies to the review on yours), then post an approved draft

• python-patterns

  Python patterns for typing, project layout, async, errors, and packaging. Load before any edit to a .py file or pyproject.toml, however small, and for Python questions.

• react-patterns

  React patterns for hooks rules, component design, and rendering performance. Load before any edit to a .jsx or .tsx file, however small, and for React questions.

• report-format

  Load before writing an audit, review, or dependency report.

• security-patterns

  Security checks for XSS, injection, auth, sessions, CSRF, CSP, secrets, and CVEs. Load before writing or reviewing code that handles auth, secrets, or user input, however small.

• slop-patterns

  How to rewrite text a slop guard flagged, and how to keep AI tells and empty filler out of prose, commit messages, PR text, code comments, docs, chat replies, and UI. Load when a guard flags AI tells, or before drafting a commit message, PR text, a README, docs, or a landing page.

• storybook-patterns

  Storybook patterns for CSF 3, args and fn() spies, play functions, story ids, and the a11y addon. Load before any edit to a *.stories.* file or .storybook/, however small, and for Storybook questions.

• tailwind-patterns

  Tailwind CSS v4 patterns for @theme config, v3-to-v4 changes, and variants. Load before any edit to a class list, stylesheet, or tailwind.config.*, however small, and for Tailwind questions.

• test-patterns

  Testing patterns for test design, fixtures, mocking, async, and coverage. Load before any edit to a test, test helper, fixture, or test config, however small, in any stack, and for testing questions.

• typescript-patterns

  TypeScript patterns for strictness, any vs unknown, satisfies, unions, generics, and module boundaries. Load before any edit to a .ts, .tsx, or tsconfig file, however small, or for TS questions.

• unslop

  Find and rewrite AI tells and filler in a file, a diff, or pasted text, then show what changed. Use when the user asks to unslop, clean up, or de-AI a piece of text.

• vps-provisioning

  VPS patterns for setup, firewall, nginx reverse proxy, Let's Encrypt, systemd, and hardening. Load before any edit to a playbook, provisioning script, or nginx or systemd config, however small.

• vue-patterns

  Vue 3 patterns for the Composition API, reactivity, and single-file components. Load before any edit to a .vue file or composable, however small, and for Vue questions.

• wcag-audit

  WCAG 2.2 AA audit checklist and severity rubric. Use for accessibility, a11y, ARIA, keyboard, focus, contrast, or form-label questions, or an accessibility audit, in any stack.

• write

  Draft outward-facing text (commit message, release notes, devnote, explainer, stakeholder summary). Use when the user asks for one of these.

Subagents (7)

• auditor

  Read-only audit of a code surface for accessibility, performance, technical debt, or documentation drift. The invoking /audit kind supplies the procedure and the report format; this shell only fixes the boundary. Not for applying fixes.

• claim-checker

  Checks the factual claims in a text before anyone relies on it, the way a reviewer who trusts nothing would. A hook asks for it before a scratch report is written or a post is sent, and names the text's claim-check id. Read-only; reports each claim nothing confirms, with the check that would settle

• debugger

  Investigates unexpected behavior to localize a fault, using read access plus at most one targeted probe edit. Never applies the fix. Use to find where and why something breaks; hand the fix back to the caller.

• gatherer

  Read-only fact gathering for one question from the investigate skill's gatherer list (precedent, tests, history). Returns facts with file:line, never opinions or advice. Use when skills/investigate names a gatherer question.

• plan-critic

  Adversarially reviews a plan file (plan mode output) against the real repo, not only the plan's own internal consistency. Checks that cited precedent says what the plan claims, catches design-integrity non-answers, and checks whether named tests would catch the failures they claim to. Use after a pl

• researcher

  Read-only external research in library docs via Context7 and web pages via WebFetch. Delegate version checks, API verification, and docs lookups here to keep network access out of code-exploring agents. Returns findings with sources; never edits, never runs shell commands.

• tester

  Adds or updates tests for recent implementation work, then runs them. Never alters implementation code to make a test pass. Use when a change needs test coverage or existing tests need extending.

Hooks (18)

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook gate-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook bash-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook guard-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook guard-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook handback-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook plan-dispatch

• PreToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook connector-dispatch

• PostToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook post-edit-dispatch

• PostToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook bash-edits

• PostToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook log-skills

• PostToolUse

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook plan-mode-context

• SessionStart

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook inject-context

• UserPromptExpansion

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook log-skills

• UserPromptSubmit

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook prompt-dispatch

• Stop

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook stop-dispatch

• SubagentStop

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook review-checks

• SubagentStop

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook subagent-dispatch

• SubagentStart

  "${CLAUDE_PLUGIN_ROOT}/bin/kit" hook inject-subagent-context