Comp AI

Apply the project’s core rules when changing TypeScript code: use Bun for package tasks, prefer the design system, validate external data without unsafe types, and follow conventions for data fetching and UI state. Run type and lint checks after changes.

中文简介正在准备中。当前显示英语简介。

• Add a TypeScript feature using project conventions
• Validate data returned by an external service
• Choose the project’s data fetching pattern

Design clearer prompts for AI tasks by setting success criteria, specifying output formats, adding examples, organizing context with XML tags, and assigning a role. The guidance helps you draft and refine instructions for more consistent responses.

中文简介正在准备中。当前显示英语简介。

• Clarifying instructions and expected output for a task
• Adding examples to guide classification or formatting
• Structuring context and constraints with XML tags

Use this skill when changing subscription billing with Stripe. It guides catalog and SKU updates, checkout, local entitlements and usage records, audit events, and webhook handling that tolerates retries and out-of-order delivery.

中文简介正在准备中。当前显示英语简介。

• Adding or updating a billing SKU
• Implementing subscription checkout
• Handling billing webhooks safely

Generate a TypeScript MCP server from an OpenAPI spec with Speakeasy, exposing API operations as tools for AI assistants. The skill covers scopes, workflow configuration, authentication, and local or Docker-based ways to run the server.

中文简介正在准备中。当前显示英语简介。

• Making API operations available as assistant tools
• Generating a TypeScript MCP server from OpenAPI
• Configuring scopes and authentication for an MCP server

Build interfaces that follow Trigger.dev v4 task runs as they happen. The skill covers subscriptions to run status, metadata, tagged runs, and streamed data from backend code or React hooks, using scoped access tokens for reads and triggers.

中文简介正在准备中。当前显示英语简介。

• Showing task status and progress in a React interface
• Subscribing to runs by ID, tag, or batch
• Displaying streamed task output as it arrives

Write Trigger.dev v4 tasks with validated inputs, scheduled runs, retries, waits, and task-to-task triggers. The skill explains how to handle Result values returned by waiting for tasks and highlights patterns that the v4 SDK does not support.

中文简介正在准备中。当前显示英语简介。

• Creating a task with input validation
• Scheduling a task with a cron pattern
• Triggering one task from another and handling its result

Use this skill to define recurring Trigger.dev tasks with cron schedules, set time zones and environments, and manage schedules through code or the dashboard. It also explains schedule payloads and common conditions that prevent a task from running.

中文简介正在准备中。当前显示英语简介。

• Defining a recurring cron task
• Scheduling runs for a particular time zone
• Creating per-user schedules with external IDs

Create and maintain Cursor rules in `.mdc` files. Learn how frontmatter, glob patterns, and automatic or manual application control which guidance applies, plus ways to organize rules and troubleshoot rules that do not apply or conflict.

中文简介正在准备中。当前显示英语简介。

• Creating a focused `.mdc` rule with examples
• Choosing files for a rule with glob patterns
• Invoking a rule manually in a prompt

Use this skill to investigate failed SDK generation from an OpenAPI spec. It helps classify errors, identify likely root causes, and outline whether an overlay can address them, the spec needs changes, or a human decision is required.

中文简介正在准备中。当前显示英语简介。

• Finding the cause of a failed Speakeasy SDK run
• Sorting OpenAPI errors by likely fix strategy
• Preparing action items for a generation failure

Use the shared CheckResultsService to retrieve the latest real results from connected integration checks, such as per-user 2FA status or bucket findings. It explains how to filter resources, validate check-specific data, and shape results for a feature.

中文简介正在准备中。当前显示英语简介。

• Showing per-person 2FA results
• Displaying integration findings in a feature
• Retrieving results for a task and source

Create, apply, and validate OpenAPI overlay files to customize SDK generation without editing the source specification. Use JSONPath to target operations or schemas and configure naming, grouping, retries, pagination, headers, enum handling, or custom security.

中文简介正在准备中。当前显示英语简介。

• Creating an overlay for SDK naming or grouping
• Applying an existing overlay to an OpenAPI spec
• Validating an overlay file

Configure a Trigger.dev v4 project through `trigger.config.ts`, including task directories, runtime, retries, lifecycle hooks, and build extensions. Examples cover tools such as Prisma, Python, Playwright, and FFmpeg, plus custom extensions and environment settings.

中文简介正在准备中。当前显示英语简介。

• Setting a Trigger.dev project runtime and retries
• Adding Prisma or Python build extensions
• Configuring browser or media processing dependencies

Build React forms with React Hook Form and Zod. The skill shows how to define validation schemas, connect fields, display errors, handle submission state, and manage complex or dynamic inputs.

中文简介正在准备中。当前显示英语简介。

• Building a validated sign-up or profile form
• Adding cross-field checks such as password confirmation
• Handling selectable or dynamically repeated fields

Review specified files for design system usage, replace legacy @trycompai/ui components and lucide-react icons with available @trycompai/design-system equivalents, and run the app type check to verify the migration.

中文简介正在准备中。当前显示英语简介。

• Migrating legacy UI components
• Replacing lucide-react icons with Carbon icons
• Checking specified files for design system compliance

Prepare an isolated worktree for a feature, ticket, or bugfix by following the repository’s setup conventions for environment files, databases, dependencies, and development servers. It also explains how to avoid conflicts when multiple worktrees are active.

中文简介正在准备中。当前显示英语简介。

• Starting a feature in a new worktree
• Setting up a branch for a Linear ticket
• Running development servers across worktrees

Review old Git worktrees and their related Postgres databases, classify which appear safe to remove, and present recommendations before cleanup. The process checks branch and pull request status, local changes, and unpushed commits to protect work that may still be needed.

中文简介正在准备中。当前显示英语简介。

• Review old worktrees before cleanup
• Find orphaned development databases
• Check whether a worktree has uncommitted work

Guide API endpoint changes so request DTOs, runtime validation, OpenAPI documentation, and MCP tool metadata stay aligned. It covers discoverability, authentication, uploads, long-running tasks, and response types that MCP tools cannot handle.

中文简介正在准备中。当前显示英语简介。

• Adding an API endpoint with a request body
• Updating DTO validation and OpenAPI metadata
• Making an endpoint available as an MCP tool

Apply consistent TypeScript and React practices for safer data handling, component structure, state, and file organization. The guidance also covers readable code, accessible interactions and forms, event naming, and user feedback with Sonner toasts.

中文简介正在准备中。当前显示英语简介。

• Writing TypeScript without unsafe casts
• Structuring React components and route files
• Reviewing state and effect usage

Set up Better Auth for TypeScript applications, including server and client configuration, database adapters, sessions, email and social sign-in, and plugins. The guide also highlights environment variables, schema generation, and security settings to review.

中文简介正在准备中。当前显示英语简介。

• Configuring Better Auth in a TypeScript app
• Connecting an ORM or database adapter
• Setting up sessions and cookie storage

Review changed code after implementation for readability, repeated logic, inconsistent patterns, unsafe boundary handling, redundant casts, and dead code. The checklist guides cleanup and calls for a TypeScript type check after fixes.

中文简介正在准备中。当前显示英语简介。

• Reviewing files after implementing a feature
• Cleaning up code after a bug fix
• Checking changed files before committing

Use these guidelines for visible UI changes to plan layouts across mobile, tablet, desktop, and large screens. They cover mobile-first breakpoints, design-system components, overflow, touch targets, long content, and checks at four screen widths.

中文简介正在准备中。当前显示英语简介。

• Building a responsive page or component
• Updating a table, toolbar, or form
• Checking a layout at several screen widths

Review security-sensitive code for serious flaws such as broken access control, tenant isolation errors, injection, exposed secrets, SSRF, and unsafe file handling. Confirm high-severity findings, fix them, and report remaining issues with their impact.

中文简介正在准备中。当前显示英语简介。

• Reviewing an API or authentication change
• Checking database queries for access control flaws
• Reviewing upload or webhook handlers

Structure data fetching so server pages provide initial data to client components, while SWR refreshes it and handles mutations through an API client. The guidance covers cache keys, server and client responsibilities, and URL based organization context.

中文简介正在准备中。当前显示英语简介。

• Building pages that render with server-fetched initial data
• Creating SWR hooks for API reads and mutations
• Keeping cached data separate by organization

Review specified files for outdated hook and API patterns, then replace them with the prescribed SWR hooks, API clients, or server API calls. The guidance covers data fetching, mutations, cache updates, and response handling, followed by a typecheck command.

中文简介正在准备中。当前显示英语简介。

• Auditing files for disallowed API and hook patterns
• Replacing raw fetch calls with the API client
• Migrating server actions and database mutations to API hooks