feat(agents): add symmetric cross-system handoff to GitHub Backlog Manager

[AhmedMustafa249]

Description

Added a symmetric 🚀 Continue with Backlog handoff to the memory agent targeting GitHub Backlog Manager, completing the round-trip between session persistence and backlog management workflows. Previously, the memory agent only surfaced an RPI continuation path; users restoring a backlog session had no direct return route. This PR closes that gap by adding the handoff entry, creating the matching github-suggest return prompt, updating both documentation guides with a Session Persistence section, and regenerating the affected plugin outputs.

Agent and Prompt

• Added 🚀 Continue with Backlog handoff to .github/agents/hve-core/memory.agent.md frontmatter with agent: GitHub Backlog Manager, prompt: "/github-suggest", and send: true, making the handoff symmetric with the existing RPI continuation path.
• Created .github/prompts/github/github-suggest.prompt.md — a new return prompt that inspects restored session context, identifies the last completed backlog pipeline stage, summarizes completed work, and proposes the next workflow step with a ready-to-use prompt.

Collections and Plugins

• Added the github-suggest prompt entry to collections/github.collection.yml (13 items) and collections/hve-core-all.collection.yml.

• Regenerated plugins/github/ and plugins/hve-core-all/ outputs; both now include github-suggest.

  plugins/hve-core/agents/memory.md was regenerated in a prior commit on this branch.

Documentation

• Inserted a ## Session Persistence section into docs/rpi/using-together.md (before ## RPI Agent: When Simplicity Fits) documenting the 💾 Save → /checkpoint continue → 🚀 Continue with RPI round-trip.
• Inserted a ## Session Persistence section into docs/agents/github-backlog/using-together.md (before ## End-to-End Walkthrough) documenting the Save → /checkpoint continue → 🚀 Continue with Backlog round-trip.
• Updated docs/agents/README.md: replaced Planned in the Utility row with [Memory Agent](github-backlog/using-together.md#session-persistence) and updated the ## Utility section body to accurately describe the Memory agent's purpose.

Related Issue(s)

Closes #506

Type of Change

Select all that apply:

Code & Documentation:

• Bug fix (non-breaking change fixing an issue)
• New feature (non-breaking change adding functionality)
• Breaking change (fix or feature causing existing functionality to change)
• Documentation update

Infrastructure & Configuration:

• GitHub Actions workflow
• Linting configuration (markdown, PowerShell, etc.)
• Security configuration
• DevContainer configuration
• Dependency update

AI Artifacts:

• Reviewed contribution with prompt-builder agent and addressed all feedback
• Copilot instructions (.github/instructions/*.instructions.md)
• Copilot prompt (.github/prompts/*.prompt.md)
• Copilot agent (.github/agents/*.agent.md)
• Copilot skill (.github/skills/*/SKILL.md)

Other:

• Script/automation (.ps1, .sh, .py)
• Other (please describe):

Sample Prompts (for AI Artifact Contributions)

User Request:

"Save my current session and then continue working on the backlog."

Execution Flow:

1. User clicks 💾 Save in the Github Backlog Manager Agent chat which hands it off to the Memory agent — a memory file is written to .copilot-tracking/memory/YYYY-MM-DD/<description>-memory.md.
2. User clicks 🚀 Continue with Backlog — the github-suggest prompt activates GitHub Backlog Manager.
3. GitHub Backlog Manager inspects conversation history and any referenced memory files.
4. Agent identifies the last completed pipeline stage (Discovery, Triage, Sprint Planning, or Execution).
5. Agent summarizes what was completed and what planning artifacts exist, then proposes the next workflow step with a ready-to-use prompt.

Output Artifacts:

• Memory file at .copilot-tracking/memory/YYYY-MM-DD/<description>-memory.md (from Memory Agent save).
• GitHub Backlog Manager session with a proposed next-step prompt based on restored context.

Success Indicators:

• GitHub Backlog Manager correctly identifies the last completed stage and proposes a contextually relevant continuation prompt.
• If no prior backlog context is found, the agent recommends starting with Discovery and provides a sample prompt.

Testing

• Verified the 🚀 Continue with Backlog button is present in the Memory Agent toolbar.
• Started a Discovery session with GitHub Backlog Manager, then clicked Save on the GitHub Backlog Manager toolbar — the session handed off to the Memory Agent, which saved the session file correctly.
• Identified a gap: the Memory Agent sometimes attempts to locate the related session file directly in .copilot-tracking/memory/, which can steer it toward the .copilot-tracking/github-issues/ folder created by the GitHub Backlog Manager instead of the correct memory file.
• Ended the Memory Agent session and clicked 🚀 Continue with Backlog — the Memory Agent handed the task back to the GitHub Backlog Manager to continue.
• Invoked /github-suggest without a prior discovery session file present — the agent correctly recommended starting with Discovery and provided a sample prompt.

Note

The session-file resolution gap noted above is a known limitation for a follow-up; it does not block the core handoff flow.

Checklist

Required Checks

• Documentation is updated (if applicable)
• Files follow existing naming conventions
• Changes are backwards compatible (if applicable)
• Tests added for new functionality (if applicable) (N/A — no test coverage surface for agent frontmatter)

AI Artifact Contributions

• Used /prompt-analyze to review contribution
• Addressed all feedback from prompt-builder review
• Verified contribution follows common standards and type-specific requirements

Required Automated Checks

• Markdown linting: npm run lint:md
• Spell checking: npm run spell-check
• Frontmatter validation: npm run lint:frontmatter
• Skill structure validation: npm run validate:skills
• Link validation: npm run lint:md-links
• PowerShell analysis: npm run lint:ps

Security Considerations

• This PR does not contain any sensitive or NDA information
• Any new dependencies have been reviewed for security issues (N/A — no new dependencies)
• Security-related scripts follow the principle of least privilege (N/A — no security scripts modified)