openspec-apply-change

Implements tasks from an OpenSpec change with planning, verification, and PR creation.

Updated Jun 11, 2026
One-click install
npx skills add https://github.com/Marcel-Carrillo/ProyectosIA --skill openspec-apply-change-marcel-carrillo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: openspec-apply-change
Source: https://github.com/Marcel-Carrillo/ProyectosIA/tree/main/.cursor/skills/openspec-apply-change
Command: npx skills add https://github.com/Marcel-Carrillo/ProyectosIA --skill openspec-apply-change-marcel-carrillo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning an approved OpenSpec change proposal into working code requires coordinating context files, task tracking, testing, and delivery. This Skill automates that implementation loop so tasks are executed, verified, and committed without manual orchestration. ## Core Features & Use Cases - Change Selection & Status Tracking: Resolves the target change by name or context, reads openspec status and openspec instructions apply JSON output, and reports progress as N/M tasks complete. - Mandatory Planning & Verification Gates: Enforces workspace isolation on a feature branch, per-layer backend/frontend planning subagents, lint preflight, and evidence-based task completion with tests and reports. - Delivery Automation: Runs adversarial review, commits with Conventional Commits, opens a pull request via gh pr create, and transitions linked Jira tickets to "En curso" and "En revisión". - Use Case: A developer asks to continue implementing the add-supplier-orders change; the Skill loads the proposal, design, and tasks, implements each pending task with tests, checks off tasks.md, and finishes with a PR. ## Quick Start Ask the agent to implement the tasks for your OpenSpec change, optionally naming the change, for example: implement the remaining tasks for the add-supplier-orders change.

Frequently Asked Questions about openspec-apply-change

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement tasks from an OpenSpec change?▼

Invoke the skill with a change name or let it infer one from context. It runs openspec status and openspec instructions apply to load context files, then works through each pending task, updating tasks.md as items complete.

What happens if the change name is ambiguous or missing?▼

The skill runs openspec list --json to retrieve available changes and prompts you to select one. If only one active change exists, it auto-selects it and announces which change is being used.

Does this skill work without the openspec CLI installed?▼

No, the openspec CLI is a stated compatibility requirement. The skill depends on commands like openspec status, openspec list, and openspec instructions apply to discover schema, context files, and task state.

Can it implement a change when artifacts are blocked or incomplete?▼

If the apply instructions return a blocked state due to missing artifacts, the skill shows the message and suggests using openspec-continue-change instead of implementing. It only proceeds when tasks are available.

How are tasks marked complete in tasks.md?▼

Tasks are checked off on disk immediately after each sub-task finishes, never batched at session end. Verification sub-tasks require evidence such as passing tests, curl checks, E2E runs, and reports before being marked complete.

Does the skill create a pull request automatically?▼

Yes, after all tasks complete and an optional adversarial review passes, it commits with a Conventional Commit message, pushes the feature branch, and runs gh pr create, then reports the PR URL.