Run agentic maintenance as local Copilot skills

[kotlarmilos]

Description

This moves the ML.NET CI scanner, scanner feedback, issue triage, and repository health agents from scheduled GitHub Agentic Workflows into repository-local Copilot skills under .github/skills/. Local execution uses the operator's authenticated tools and keeps GitHub writes behind explicit approval while preserving the existing playbooks as progressive-disclosure references. This removes scheduled execution and compiled workflow locks until repository billing and authentication are available. The CI authentication implementation in #7652 remains the reference for a later follow-up.

[Copilot]

Pull request overview

This PR migrates ML.NET’s agentic maintenance automation from scheduled GitHub Agentic Workflows into repository-local Copilot “skills” under .github/skills/, intentionally removing scheduled execution and compiled workflow lock files until billing/auth is available.

Changes:

• Removes scheduled/dispatch GitHub Agentic Workflow artifacts (compiled *.lock.yml, agent prompt markdown, and the AW action lock).
• Adds repository-local Copilot skills (SKILL.md + playbook references) for repo health (check/investigate/groom), CI scanning (scan + feedback), and issue triage.
• Updates .github/copilot-instructions.md to document the new local-skill entry points.

Show a summary per file

File	Description
.github/workflows/repo-health-investigate.lock.yml	Removes compiled workflow lock for repo-health investigation (disables scheduled/dispatch execution path).
.github/workflows/repo-health-groom.lock.yml	Removes compiled workflow lock for dashboard grooming (disables scheduled execution path).
.github/workflows/issue-triage.agent.md	Removes workflow-based issue triage agent definition (superseded by local skill).
.github/workflows/ci-scan.agent.md	Removes workflow-based CI scan agent definition (superseded by local skill).
.github/workflows/ci-scan-feedback.agent.md	Removes workflow-based CI scan feedback agent definition (superseded by local skill).
.github/skills/repo-health-investigate/SKILL.md	Adds local skill entry point and guardrails for investigating a single repo-health finding.
.github/skills/repo-health-investigate/references/playbook.md	Reframes investigation playbook for local execution + explicit approval before writes.
.github/skills/repo-health-groom/SKILL.md	Adds local skill entry point and rules for conservative dashboard grooming.
.github/skills/repo-health-groom/references/playbook.md	Reframes grooming playbook for local execution + explicit approval before writes.
.github/skills/repo-health-check/SKILL.md	Adds local skill entry point for running repo-health checks and drafting dashboard updates.
.github/skills/repo-health-check/references/playbook.md	Reframes orchestrator playbook for local execution and local investigation routing.
.github/skills/issue-triage/SKILL.md	Adds local skill for issue classification and drafting labels/comment.
.github/skills/ci-scan/SKILL.md	Adds local skill for CI failure scanning and drafting KBE issues (approval-gated).
.github/skills/ci-scan/references/playbook.md	Adds local CI-scan playbook aligned with the former workflow methodology.
.github/skills/ci-scan/references/ci-scan.instructions.md	Adds local shared methodology and updates paths/wording for local drafts vs workflow writes.
.github/skills/ci-scan-feedback/SKILL.md	Adds local skill for auditing CI-scan quality and proposing improvements (read-only).
.github/copilot-instructions.md	Documents the new repository-local skills and clarifies approval-gated writes.
.github/aw/actions-lock.json	Removes gh-aw action lockfile (no longer needed after moving away from scheduled AW workflows).

Copilot's findings

• Files reviewed: 22/22 changed files
• Comments generated: 0