claude

TDD task breakdown and TASKS.md management. Use when planning a new feature, breaking down work into implementation tasks, creating or updating TASKS.md, estimating effort, or organizing multi-step work into phases. Also use when translating a plan document (like docs/plans/*.md) into TASKS.md — includes systematic extraction checklist to ensure nothing is missed. Trigger on: feature breakdown, task planning, "write tasks for", "create TASKS.md", plan-to-tasks conversion, sprint planning, implementation ordering, or when the user mentions wanting to organize implementation steps for a feature.

MCP integration for kendo project management. Use when creating, updating, moving, or searching issues via MCP tools. Also use when writing user stories, logging time, managing epics, linking branches to issues, or triaging reports (bug reports, feedback). Provides access to projects, issues, lanes, sprints, epics, reports, and teams through MCP resources and tools.

Triage incoming reports (bug reports, feedback, feature requests) from the Kendo report queue. Fetches pending reports via MCP, cross-references with existing open issues to find overlaps, then walks the user through each report one-by-one using AskUserQuestion for a verdict: promote to issue, dismiss, or combine with another report. Promoted reports get proper user stories or bug reports following project conventions. Use this skill whenever the user mentions "triage", "triage reports", "check reports", "process feedback", "review reports", "incoming reports", "report queue", "what reports do we have", or any variant of wanting to process the Kendo report queue. Also trigger when the user asks about unprocessed feedback or pending bug reports.

Prepare a kendo issue for development: assign to current user, create a feature branch, link it, move to In Progress, and optionally check out in a worktree with full setup. Usage: /prepare-issue <issue_key_or_id> [N]. Use whenever the user wants to start working on an issue, prepare an issue, pick up an issue, or says "prepare issue", "start issue", "work on issue". Also trigger when combining issue assignment with branch creation, or when the user wants to set up a worktree for an issue.

Load the full context of the current branch and check alignment with the base: git diff summary, commit history, plan/task files, decisions, and (if the project has an issue-tracker skill) the upstream issue details. Reports how far behind the base the branch is and offers to merge on confirmation. Resolves merge conflicts intelligently (auto-resolves obvious ones, asks on ambiguous ones). Use this skill whenever the user wants to understand what the current branch is about, needs context after a /clear, starts a new session, says "what am I working on", "load context", "catch me up", "branch context", "where was I", "summarize this branch", "catchup", "sync my branch", "align with base", "update branch", or any variant of wanting to understand or sync the current state of work. Also use when the user just passes a branch name or issue key and expects you to load its context. Trigger on session start when the branch is not the default branch and the user's first message implies they want to resume work.

Synchronize the kendo board with GitHub branch and PR status. Use this skill whenever the user wants to check board alignment, sync the board, see what issues are stale, find merged PRs that haven't been moved to Done, or get an overview of branch status for open issues. Trigger on phrases like "sync board", "check board", "board status", "align board", "update board", "which issues are done", or "check branches".

Structured bug-fix workflow: identify the issue, reproduce the defect, confirm it isn't already fixed on the base branch, capture the investigation in docs/bugs/<slug>/BUG.md, propose the fix and wait for developer approval, implement it, and gate on the bug-fix-verifier agent before PR. Use whenever the user wants to fix a bug, says "fix bug", "bug fix", "debug this issue", "work on bug <key>", "/fix-bug", or is picking up a bug-type issue. Prefer this over /plan-feature for bugs — bugs don't have wireframes or acceptance criteria, they have "does the defect still reproduce?". This skill is the bug-side analogue of /plan-feature + /implement-plan collapsed into one lighter flow.

Create a pull request for the current branch, with optional issue feedback (when the branch links to a tracked issue) and optional review handoff (when a fresh review file exists in the plan directory). Use this skill whenever the user wants to create a PR, open a pull request, submit their work for review, or is done with a feature/fix branch. Also triggers on phrases like "make a PR", "open PR", "create pull request", "submit for review", "I'm done with this branch", or "push and PR".

Multi-agent review of the current branch vs the base branch. Spawns acceptance + simplicity (+ conditional silent-failure / efficiency) reviewers in parallel and synthesizes the output into `<plan-directory>/REVIEW_CLAUDE.md` for `/pr` to consume. Use whenever the user wants to review a branch, says "review my changes", "review branch", "check my code", "what did I change", "write the handoff", "branch review", "review handoff", or is about to open a PR and wants the review artifact first. **Writes a file** — if there's no plan directory, falls back to a lightweight chat-only review and nothing is written.

Structured feature planning workflow: interrogate the developer with hard, codebase-informed questions until requirements are crystal clear, then produce PLAN.md and DECISIONS.md. Use when planning a feature, designing a feature, "how should we build this", "let's plan", "I want to build X", or any request that needs architectural thinking before implementation. If a feature touches multiple layers (frontend + backend + migration), trigger this skill before coding.

Create a new branch from `{{DEFAULT_BRANCH}}`, named after a Kendo issue (e.g. `{{ISSUE_KEY_PREFIX}}-0141-short-description`). Use whenever the user wants to start a new feature, create a branch, begin work on an issue, or says "new branch".

Kendo CLI tool for managing issues, projects, sprints, epics, reports, and time tracking directly from the terminal. Use this skill whenever interacting with kendo issues, checking board status, creating or updating issues, searching across projects, logging time, managing sprints and epics, or triaging reports (bug reports, feedback) — especially when the MCP server is unavailable or when a quick CLI command is simpler than an MCP tool call. Also use when the user mentions "kendo cli", "kendo command", or wants to run kendo operations via the terminal.

Implement a feature directly from PLAN.md and DECISIONS.md without a TASKS.md breakdown. Use this for smaller plans where the plan's Approach section is already a good enough roadmap and a separate task file would be overkill. Loads the full plan context (goal, decisions, scope, approach, acceptance criteria, wireframes), executes the work end-to-end with TDD discipline, runs verification and the acceptance-reviewer agent, then captures learnings in IMPLEMENTATION.md so future sessions can resume. Trigger on: "implement the plan", "implement plan", "build this plan", "execute the plan", "/implement-plan", "ship this plan", "just do the plan", or whenever the user wants to build a feature straight from a plan document without writing TASKS.md first. Prefer /next when TASKS.md already exists in the plan directory.

Continue working through TASKS.md — find the next uncompleted task, execute it with TDD flow, and mark it done. Use this skill whenever the user wants to continue implementation work, pick up where they left off, resume after /clear, or work through a task list. Trigger on: "next task", "continue working", "keep going", "what's next", "pick up where I left off", "resume work", "next step", or any variant of wanting to progress through planned tasks.

Generate structured UI wireframe specifications (WIREFRAMES.md) for feature plans. Creates AI-readable screen specs with ASCII wireframes, design tokens, component breakdowns, and interaction specs. Use after /plan-feature produces PLAN.md, whenever the user mentions "wireframe", "create wireframes", "design the screens", "screen specs", "UI spec", or wants to define what frontend pages/components should look like before implementation. Also trigger when the user says "what should the page look like" or "define the UI" for a planned feature.

Watch a PR's GitHub Actions CI, automatically fix failures, and push until green. Use this skill whenever the user says "babysit", "watch ci", "fix ci", "make ci green", "monitor the PR", "keep pushing until green", or after creating a PR when they want hands-off CI resolution. Also use proactively after /commit or /pr when CI is likely to run and the user seems done coding.

Create small, focused commits with proper messages and push to remote. Use whenever the user wants to commit, save progress, push changes, or says "commit", "push", or "save my work".

Nightwatch error and performance monitoring. Use when checking production or staging errors, triaging exceptions, investigating slow routes, reviewing application health, or managing issue status (resolve/ignore). Also use when the user mentions Nightwatch, production errors, exceptions, slow endpoints, or application monitoring. ALWAYS use this skill before calling any mcp__nightwatch__ tool.

Release a new version of a Go CLI tool by tagging and pushing to trigger GoReleaser. Use whenever the user wants to release the CLI, tag a CLI version, publish the CLI, cut a CLI release, or says "release cli", "tag cli", "new cli version". Also use when the user asks about the CLI release process or how to ship a new CLI version.

Run a research query, produce a structured markdown report, and file it so knowledge accumulates instead of dying in conversations. Use when the user asks to research a topic, investigate a question, compare options, analyze competitors, or any exploratory query where the output should be preserved. Trigger on: "research", "investigate", "deep dive", "compare", "analyze", "look into", "what do we know about", or any question that deserves a filed answer rather than a chat response.

Write a retrospective capturing what went wrong, why, and what changed. Use when something went wrong that's worth documenting, after a failed approach, or when the user says "retro", "retrospective", "lessons learned", "post-mortem", or "what went wrong". Also trigger proactively when a significant mistake or repeated issue is detected in the conversation.

Review and improve MCP tool and resource descriptions for Tool Search discoverability. Use after adding or modifying MCP tools or resources to ensure descriptions follow best practices. Trigger on: "review mcp descriptions", "check tool descriptions", "optimize mcp tools", or after creating new MCP tools/resources.

Full project setup for Laravel + Vue + MinIO projects: runs backend, frontend, and MCP in parallel. Supports multiple worktrees with automatic port assignment.

What this skill does and when Claude should use it. Be specific and slightly "pushy" — Claude tends to undertrigger skills, so include contexts where it should activate even if the user doesn't explicitly ask for it. Example: "Generates changelog entries from git history. Use this whenever the user mentions changelogs, release notes, version bumps, or asks what changed between two commits."