backlog-execute

Creates and updates work items in Azure DevOps, GitHub, and Jira from reviewed handoff files.

1.4k|284|Updated Nov 2, 2025
One-click install
npx skills add https://github.com/microsoft/hve-core --skill backlog-execute
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: backlog-execute
Source: https://github.com/microsoft/hve-core/tree/main/.github/skills/project-planning/backlog-execute
Command: npx skills add https://github.com/microsoft/hve-core --skill backlog-execute

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Applying planned backlog changes to a live tracker is error-prone: wrong platforms, duplicated creates after interruptions, unreviewed plans, and leaked placeholder IDs. This Skill executes mutating backlog operations against Azure DevOps, GitHub, or Jira with enforced safety gates so every write is confirmed, logged, and resumable.

Core Features & Use Cases

  • Single-item creation: Guided conversational creation of one work item with platform-validated types, hierarchy checks, and field collection.
  • Handoff execution: Processes a reviewed handoff file into ordered create, update, link, comment, and close operations with per-operation logging to handoff-logs.md.
  • Safety protocols: Three-tier autonomy gating, dry-run rendering, resumable execution from logs, human-review checkbox enforcement, and content sanitization before any tracker call.
  • Use Case: A planner produces a handoff file for a sprint's worth of GitHub issues; after human review, run this Skill to create the issues in dependency order, and if the run is interrupted, resume without duplicating completed creates.

Quick Start

Run the backlog-execute skill to apply the reviewed handoff file at .copilot-tracking/handoff.md to our GitHub repository with partial autonomy.

Frequently Asked Questions about backlog-execute

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

FAQPage Schema
How do I create a work item in Azure DevOps, GitHub, or Jira from an AI assistant?

Invoke the skill in add mode with the item description; it resolves the platform, confirms the destination, collects fields conversationally, validates the hierarchy, and creates the item. The returned item key is logged before the run ends.

How do I apply a backlog handoff file to a tracker without duplicating items?

Run the skill in run mode against the handoff file. It rebuilds the temporary-identifier mapping from successful live entries in handoff-logs.md and resumes from the first operation without a successful entry, so completed creates are never re-run.

Does backlog-execute support dry runs before mutating a tracker?

Yes. The --dry-run flag resolves and validates the full operation set and renders exactly what would be sent, without making any mutating call. Results are written to handoff-dryrun.md, never to the live log.

What permissions does backlog-execute need for Jira?

Jira execution requires JIRA_BASE_URL plus either JIRA_API_TOKEN or JIRA_PAT, and terminal access for the Jira CLI. Azure DevOps and GitHub require their respective write tool families available in the active context.

Why does backlog-execute stop before processing my handoff file?

Processing halts when the handoff contains an unchecked human-review checkbox, the platform or destination is unconfirmed, or no compatible write surface exists in the current context. The skill reports the artifact path and the specific blocking item.

When should I use backlog-plan instead of backlog-execute?

Use backlog-plan for discovery, triage, sprint planning, and any read-only analysis; it produces the handoff file. Use backlog-execute only for mutating operations after that handoff has been reviewed by a human.