Fix: Require explicit assignee verification in auto-assign workflow

[Copilot]

The auto-assign workflow was assigning issues that already had assignees. The AI agent correctly filtered search results with no:assignee but failed to re-verify the assignees field after calling issue_read, creating a race condition window.

Changes

Strengthened prompt verification requirements:

• Added explicit two-step verification mandate (search filter + post-read check)
• Inserted numbered 6-step verification checklist with CRITICAL designation
• Added explicit EXIT condition: "IF assignees is NOT empty, EXIT WITHOUT ASSIGNING"
• Changed conditional from "If you find a match" to "If you find a match AND the issue has no assignees"

Before:

- **Has no assignees** - When you retrieve issues from GitHub, explicitly 
  check the `assignees` field. Skip any issue where `issue.assignees` is 
  not empty or has length > 0.

After:

- **Has no assignees** - This is CRITICAL. You MUST verify this in two places:
  1. Use `no:assignee` in your search query
  2. After retrieving the issue details, explicitly check that `issue.assignees` 
     is an empty array or has length 0
  3. **SKIP AND EXIT if `issue.assignees` is not empty**

**Verification Steps (REQUIRED):**
1. Search for issues using `no:assignee` filter
2. Pick the oldest unassigned issue from search results
3. Retrieve full issue details using `issue_read`
4. **VERIFY** that `issue.assignees` is empty (length === 0)
5. **IF assignees is NOT empty, EXIT WITHOUT ASSIGNING**
6. Only if `issue.assignees` is empty, proceed to list contributors and assign

Recompiled workflow lock file with updated prompt.

Original prompt

Reference: https://github.com/githubnext/gh-aw/actions/runs/21310240804/job/61345358327

Debug this run and make sure the agent only assigns issues without an assignee

Custom agent used: agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.