spec-kit-command-cursor

Converts natural language ideas and requirements into structured specification documents through a Cursor IDE command interface. The toolkit prompts users to articulate project scope, requirements, and constraints, then synthesizes responses into a formatted specification that serves as the single source of truth for development. Works by intercepting the /specify command in Cursor, capturing user input through guided prompts, and formatting output as markdown specifications compatible with spec-driven development workflows.

Breaks down specifications into hierarchical development plans with phases, milestones, and dependencies. The /plan command accepts a specification document and generates a structured plan that maps requirements to implementation phases, identifies critical path items, and suggests task ordering. Implementation uses prompt-based decomposition where the toolkit guides users through planning decisions (timeline, resource constraints, risk factors) and synthesizes responses into a markdown plan document with clear phase boundaries and success criteria.

Converts development plans into granular, assignable tasks with acceptance criteria and implementation hints. The /tasks command parses a plan document and generates a task list where each item includes a clear description, acceptance criteria, estimated effort, and optional implementation notes. Works by analyzing plan phases and milestones, then prompting users to define task granularity and acceptance criteria, synthesizing responses into a structured task document that can be imported into issue trackers or used as a checklist.

Provides a shell-based command registration system that hooks into Cursor IDE's slash command interface, allowing /specify, /plan, and /tasks commands to be invoked directly from the editor. Implementation uses shell scripts that register commands with Cursor's command palette, capture user input through the editor's prompt system, and execute the toolkit's logic in-process. Commands integrate with Cursor's native UI for prompts and file creation, ensuring seamless editor experience without external windows or context-switching.

Maintains explicit references between specification sections and plan phases, enabling bidirectional navigation and impact analysis. When /plan is executed on a specification, the generated plan document includes references back to the spec sections it addresses, and plan phases are tagged with requirement IDs. This allows developers to trace any plan phase back to its originating requirement and identify which spec sections are covered by which plan phases. Implementation uses markdown link syntax and structured headers to create a queryable relationship graph without requiring a database.

Provides pre-built specification templates that guide users through defining key sections (scope, requirements, constraints, acceptance criteria) without starting from a blank page. Templates are markdown files with section headers and placeholder text that prompt users to fill in project-specific details. The /specify command can optionally use a template as a starting point, pre-populating structure and asking users to customize each section. Implementation stores templates in the toolkit directory and allows users to create custom templates by copying and modifying existing ones.

Generates well-formatted markdown documents for specifications, plans, and tasks with consistent heading hierarchy, section organization, and link syntax. The toolkit uses shell scripts to construct markdown output with proper formatting (headers, lists, code blocks, links) that renders correctly in markdown viewers and GitHub. Implementation uses printf/echo commands to build markdown strings with proper escaping and indentation, ensuring output is both human-readable and machine-parseable. All generated documents follow a consistent structure that makes them easy to navigate and version control.

Collects structured user input through a series of interactive prompts in the Cursor editor, guiding users through specification, planning, and task definition workflows. Prompts are displayed via Cursor's native input dialog system, capturing responses as text that are then processed and formatted into documents. Implementation uses shell read commands and Cursor's prompt API to create a conversational workflow where each prompt builds on previous responses, allowing users to refine their thinking as they answer questions about requirements, timeline, and constraints.