kickoff
Decomposes a larger prompt into a structured, skill-annotated task plan. Pass it a description of the work you want done and it returns a step-by-step plan where each task is mapped to the right finstreet-fe skill. Use when starting a new feature, tackling a multi-step request, or when you want Claude Code to work through a complex prompt task by task with progress tracking. Trigger this whenever the user says 'kickoff', 'plan this', 'break this down', 'structure this', or passes a long multi-part prompt they want organized.
What this skill does
# Kickoff — Prompt Decomposition & Skill Mapping You take a freeform description of work and turn it into a structured, trackable task plan. Each task is annotated with which skill(s) to invoke, so the user (or an orchestrator) can work through them one by one with full context. ## What You Do 1. **Read the user's prompt** — understand what they want built or changed 2. **Decompose into tasks** — break the work into discrete, ordered steps 3. **Map skills to tasks** — annotate each step with the skill(s) that apply 4. **Create a task list** — use Claude Code's task tools so progress is visible 5. **Present the plan** — show the user the structured plan for confirmation before starting ## How to Decompose Think about the work in terms of **what needs to exist when it's done** and work backwards: - Does it need a new API endpoint? → `secure-fetch` first - Does it need a new page? → `page` (and probably `routes` before it) - Does it need a form? → `form` (and `page` for the shell) - Does it need a loading state? → `loading` (after the page exists) - Does it need e2e tests? → `e2e-test` (after everything else) Order tasks so that each step's output feeds the next step's input. Backend before frontend. Shell before content. Data before UI. ## Skill Catalog Reference this catalog when mapping tasks to skills. Each entry describes what the skill does and when to reach for it. ### Backend & Data | Skill | What It Does | When to Use | |-------|-------------|-------------| | **secure-fetch** | Creates type-safe server/client HTTP request functions using `@finstreet/secure-fetch`. Covers schema, server.ts, and client.ts files. | Any time you need to call a backend API endpoint — GET, POST, PUT, DELETE. Always the first step when backend integration is needed. | | **mock-api** | Creates mock API endpoints that plug into the secure-fetch pattern. | When the real backend endpoint isn't ready yet and you need to unblock frontend work. | ### Pages & Routing | Skill | What It Does | When to Use | |-------|-------------|-------------| | **routes** | Resolves, adds, and edits entries in `routes.ts`. | When a new page needs a route, or you need to look up an existing route path. | | **path-resolver** | Resolves feature and backend file paths based on naming conventions. | When you need to figure out where files should be created for a given feature. Invoke directly: `/path-resolver`. | | **page** | Builds Next.js page shells — metadata, params, header selection, content wrappers. | When creating a new page. Pages are thin shells; the actual content is built by other skills (form, list, task-group, etc.). | | **loading** | Builds `loading.tsx` skeleton pages that mirror page content structure. | After a page exists and you want a loading skeleton for it. | ### Forms & Input | Skill | What It Does | When to Use | |-------|-------------|-------------| | **form** | Full form implementation using `@finstreet/forms` — schema, fields, action, default values, config, and components. | Any multi-field form: create, edit, or inquiry step forms. The most comprehensive skill. | | **simple-form** | Lightweight action-only forms without input fields (just a submit button). | Confirmation dialogs, one-click actions, or forms with no user input fields. | | **inquiry-process** | Multi-step form wizard using `@finstreet/forms` and `@finstreet/ui`. | When building a multi-page wizard (e.g., an onboarding flow with progress bar and multiple steps). | ### UI Components & Patterns | Skill | What It Does | When to Use | |-------|-------------|-------------| | **ui** | Expert guide to PandaCSS layout primitives and `@finstreet/ui` components. | When building any UI component — layout, styling, responsive design, or component composition. | | **modal** | Implements modals — Zustand store, modal component, and optional open button. | When you need a dialog/modal overlay. | | **task-group** | Builds TaskGroup patterns — TaskPanels, ActionPanels, and the TaskGroup wrapper. | When displaying a set of tasks a user must complete before proceeding (e.g., checklist-style UI). | | **list-actions** | Adds pagination, search, sorting, filtering, and grouping to an InteractiveList. | When an existing list needs server-side pagination, search, or sorting capabilities. | | **contract-upload** | Builds a contract upload page — file upload, scan status, document management, signature process start. | When building a page where users upload contract documents for the signature process. | | **document-exchange** | Builds document exchange pages — upload, download, and manage documents in collapsible request groups. | When building a page for document management with grouped document requests. | ### Testing & Quality | Skill | What It Does | When to Use | |-------|-------------|-------------| | **e2e-test** | Writes Playwright e2e tests — form modules, card CRUD, inquiry pages, happy paths. | After a feature is built and you need automated browser tests. | ### Git & Workflow | Skill | What It Does | When to Use | |-------|-------------|-------------| | **commit** | Reviews, stages, commits, and pushes changes. | When you're done with a piece of work and want to commit. Invoke: `/finstreet-dev:commit`. Requires the `finstreet-dev` plugin. | | **pr** | Creates a pull request for the current branch. | After all work is committed and you want to open a PR. Invoke: `/finstreet-dev:pr`. Requires the `finstreet-dev` plugin. | | **new-feature-branch** | Creates a git branch following Conventional Branch naming. | At the very start of a new feature, before any code changes. Invoke: `/finstreet-dev:new-feature-branch`. Requires the `finstreet-dev` plugin. | ## Task Plan Format Present the plan to the user as a numbered list. Each task should have: 1. **Title** — short description of what to do 2. **Skill** — which skill to invoke (or "manual" if no skill applies) 3. **Details** — what specifically needs to happen in this step 4. **Depends on** — which prior tasks must be complete first ### Example Given a prompt like: *"I need a new page where property managers can upload documents for a financing case. It needs a form for metadata, file upload, and a loading skeleton. The API endpoint exists already."* The plan would be: ``` ## Task Plan 1. **Resolve paths and routes** Skill: `path-resolver`, `routes` → Determine feature directory and add route to routes.ts 2. **Create page shell** Skill: `page` → Build the Next.js page with metadata, params, and sub-page header Depends on: 1 3. **Build metadata form** Skill: `form` → Schema, fields, action, default values, config, and form components Depends on: 1 4. **Build document upload section** Skill: `document-exchange` or `contract-upload` → File upload with request groups and status tracking Depends on: 1 5. **Assemble page content** Skill: `ui` → Compose form and upload section into the page content component Depends on: 2, 3, 4 6. **Create loading skeleton** Skill: `loading` → Build loading.tsx that mirrors the page structure Depends on: 5 7. **Write e2e tests** Skill: `e2e-test` → Happy path test covering form submission and file upload Depends on: 5 ``` ## Creating the Task List After presenting the plan and getting user confirmation, create tasks using Claude Code's task tools: ``` TaskCreate: task 1 title — skill: X — details TaskCreate: task 2 title — skill: Y — details (depends on: 1) ... ``` This gives the user a live progress tracker. As each task is worked on, update its status to `in_progress`, then `completed` when done. ## Rules 1. **Always present the plan before starting work** — the user should confirm or adjust before you begin 2. **Order matters** — backend before frontend, shell before content, data before UI 3. **One skill per task** where possible — if a task needs two skills, split it into two tasks 4. **Include git tasks** — if the `finstreet-dev` plugi
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.