What problem does it solve? Turning an OpenSpec change proposal into actual code requires manually tracking which tasks are done, which artifacts provide context, and what to do when blocked. This Skill automates that loop by reading the change's status and apply instructions from the openspec CLI, then working through tasks one by one. ## Core Features & Use Cases - Change Selection: Infers the target change from conversation context, auto-selects when only one active change exists, or prompts the user to pick from openspec list --json output. - Status-Aware Execution: Parses openspec status and openspec instructions apply JSON to understand the schema, progress, and remaining tasks before writing any code. - Task Loop with Guardrails: Implements each pending task, marks checkboxes complete in the tasks file, and pauses on ambiguity, design issues, or errors instead of guessing. - Use Case: A developer has an approved OpenSpec change with 7 tasks. They invoke this Skill, which reads the proposal, specs, design, and tasks files, then implements each task sequentially, reporting progress as "N/M tasks complete" until done or blocked. ## Quick Start Ask the AI to start implementing the tasks for your OpenSpec change, optionally naming the change, and it will read the context files and work through the task list.
